®
9.22
®
This document is an individual chapter from
SAS/STAT
®
9.22 User’s Guide
.
The correct bibliographic citation for the complete manual is as follows: SAS Institute Inc. 2010.
Guide . Cary, NC: SAS Institute Inc.
SAS/STAT
®
9.22 User’s
Copyright © 2010, SAS Institute Inc., Cary, NC, USA
All rights reserved. Produced in the United States of America.
For a Web download or e-book : Your use of this publication shall be governed by the terms established by the vendor at the time you acquire this publication.
U.S. Government Restricted Rights Notice : Use, duplication, or disclosure of this software and related documentation by the U.S. government is subject to the Agreement with SAS Institute and the restrictions set forth in FAR 52.227-19,
Commercial Computer Software-Restricted Rights (June 1987).
SAS Institute Inc., SAS Campus Drive, Cary, North Carolina 27513.
1st electronic book, May 2010
SAS
®
Publishing provides a complete selection of books and electronic products to help customers use SAS software to its fullest potential. For more information about our e-books, e-learning products, CDs, and hard-copy books, visit the
SAS Publishing Web site at support.sas.com/publishing or call 1-800-727-3228.
SAS
® and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute
Inc. in the USA and other countries. ® indicates USA registration.
Other brand and product names are registered trademarks or trademarks of their respective companies.
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
PROC MCMC Compared with Other SAS Procedures
. . . . . . . . . . . .
Getting Started: MCMC Procedure
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
BEGINNODATA/ENDNODATA Statements
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4132
4133
4135
4135
4136
4139
4140
. . . . . . . . . . . . . . . . . . . . . . . . 4141
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4141
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4143
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4147
. . . . . . . . . . . . . . . . . . . . . . . . . . . 4147
4102
4103
4103
4104
4112
4116
4119
4120
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tuning the Proposal Distribution
Initial Values of the Markov Chains
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
4148
4149
4150
4153
4154
4155
4166
Using Density Functions in the Programming Statements
. . . . . . . . . . . . 4167
. . . . . . . . . . . . . . . . . . . . . . . . . . .
Multivariate Density Functions
4169
. . . . . . . . . . . . . . . . . . . . . . . . . 4171
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . .
4174
4176
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4181
Regenerating Diagnostics Plots
. . . . . . . . . . . . . . . . . . . . . . . .
4182
4102
F
Chapter 52: The MCMC Procedure
Posterior Predictive Distribution
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
Floating Point Errors and Overflows
. . . . . . . . . . . . . . . . . . . . . .
4185
4190
4190
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4193
4195
4195
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4200
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4201
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
4202
Example 52.1: Simulating Samples From a Known Density
Example 52.2: Box-Cox Transformation
. . . . . . . . .
4202
. . . . . . . . . . . . . . . . . . . . 4207
Example 52.3: Generalized Linear Models
. . . . . . . . . . . . . . . . . .
Example 52.4: Nonlinear Poisson Regression Models
. . . . . . . . . . . .
Example 52.5: Random-Effects Models
Example 52.6: Change Point Models
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
4216
4229
4238
4254
Example 52.7: Exponential and Weibull Survival Analysis
. . . . . . . . . .
4258
. . . . . . . . . . . . . . . . . . . . . . . . . . . 4271
Example 52.9: Normal Regression with Interval Censoring
. . . . . . . . .
4288
Example 52.10: Constrained Analysis
. . . . . . . . . . . . . . . . . . . .
Example 52.11: Implement a New Sampling Algorithm
. . . . . . . . . . .
4290
4296
Example 52.12: Using a Transformation to Improve Mixing
. . . . . . . . . . 4307
Example 52.13: Gelman-Rubin Diagnostics
. . . . . . . . . . . . . . . . . . 4317
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4324
The MCMC procedure is a general purpose Markov chain Monte Carlo (MCMC) simulation procedure that is designed to fit Bayesian models. Bayesian statistics is different from traditional statistical methods such as frequentist or classical methods. For a short introduction to Bayesian analysis and
of difficulty.
In essence, Bayesian statistics treats parameters as unknown random variables, and it makes inferences based on the posterior distributions of the parameters. There are several advantages associated with this approach to statistical inference. Some of the advantages include its ability to use prior information and to directly answer specific scientific questions that can be easily understood. For further discussions of the relative advantages and disadvantages of Bayesian analysis, see the section
“ Bayesian Analysis: Advantages and Disadvantages ” on page 147.
It follows from Bayes’ theorem that a posterior distribution is the product of the likelihood function and the prior distribution of the parameter. In all but the simplest cases, it is very difficult to obtain
PROC MCMC Compared with Other SAS Procedures
F
4103 the posterior distribution directly and analytically. Often, Bayesian methods rely on simulations to generate sample from the desired posterior distribution and use the simulated draws to approximate the distribution and to make all of the inferences.
PROC MCMC is a flexible simulation-based procedure that is suitable for fitting a wide range of
Bayesian models. To use the procedure, you need to specify a likelihood function for the data and a prior distribution for the parameters. You might also need to specify hyperprior distributions if you are fitting hierarchical models. PROC MCMC then obtains samples from the corresponding posterior distributions, produces summary and diagnostic statistics, and saves the posterior samples in an output data set that can be used for further analysis. You can analyze data that have any likelihood, prior, or hyperprior with PROC MCMC, as long as these functions are programmable using the SAS
DATA step functions. The parameters can enter the model linearly or in any nonlinear functional form. The default algorithm that PROC MCMC uses is an adaptive blocked random walk Metropolis algorithm that uses a normal proposal distribution.
PROC MCMC is unlike most other SAS/STAT procedures in that the nature of the statistical inference is Bayesian. You specify prior distributions for the parameters with the likelihood function for the data with
statements and statements. The procedure derives inferences from simulation rather than through analytic or numerical methods. You should expect slightly different answers from each run for the same problem, unless the same random number seed is used. The model specification is similar to PROC NLIN, and PROC MCMC shares much of the syntax of
PROC NLMIXED.
Note that you can also carry out a Bayesian analysis with the GENMOD, PHREG, and LIFEREG procedures for generalized linear models, accelerated life failure models, Cox regression models, and piecewise constant baseline hazard models (also known as piecewise exponential models). See
Chapter 37, “ The GENMOD Procedure ,” Chapter 64, “ The PHREG Procedure ,” and Chapter 48,
There are three examples in this “Getting Started” section: a simple linear regression, the Behrens-
Fisher estimation problem, and a random effects model. The regression model is chosen for its simplicity; the Behrens-Fisher problem illustrates some advantages of the Bayesian approach; and the random effects model is one of the most prevalently used models.
Keep in mind that
statements declare the parameters in the model,
statements declare the prior distributions, and
statements declare the likelihood for the data. In most cases, you do not need to supply initial values. The procedure advises you if it is unable to generate starting values for the Markov chain.
4104
F
Chapter 52: The MCMC Procedure
This section illustrates some basic features of PROC MCMC by using a linear regression model. The model is as follows:
Y i
D ˇ
0
C ˇ
1
X i
C i for the observations i D 1; 2; : : : ; n
.
The following statements create a SAS data set with measurements of
Height and
Weight for a group of children:
title 'Simple Linear Regression'; data Class; input Name $ Height Weight @@; datalines;
Alfred 69.0 112.5
Carol
Jane
John
Louise
Robert
62.8 102.5
59.8
59.0
56.3
84.5
99.5
77.0
64.8 128.0
William 66.5 112.0
Alice
Henry
Janet
Joyce
Mary
56.5
51.3
84.0
63.5 102.5
62.5 112.5
50.5
66.5 112.0
Ronald 67.0 133.0
;
Barbara 65.3
James 57.3
Jeffrey 62.5
Judy 64.3
Philip
Thomas
98.0
83.0
84.0
90.0
72.0 150.0
57.5
85.0
The equation of interest is as follows:
Weight i
D
ˇ
0
C
ˇ
1
Height i
C i
The observation errors, i
, are assumed to be independent and identically distributed with a normal distribution with mean zero and variance
2
.
Weight i normal
.ˇ
0
C ˇ
1
Height i
;
2
/
The likelihood function for each of the follows:
Weight
, which is specified in the
statement, is as p.
Weight j ˇ
0
; ˇ
1
;
2
; Height i
/ D .ˇ
0
C ˇ
1
Height i
;
2
/ where p.
j / denotes a conditional probability density and parameters in the likelihood:
ˇ
0
,
ˇ
1
, and
2
. You use the is the normal density. There are three
statement to indicate that these are the parameters in the model.
Suppose that you want to use the following three prior distributions on each of the parameters:
.ˇ
0
/
.ˇ
1
/
.
2
/
D
D
D
.0; var
D 1e6/
.0; var D 1e6/ f i
.
shape
D
3=10; scale
D
10=3/ where .
/ indicates a prior distribution and distribution. The normal priors on
ˇ
0 and f i is the density function for the inverse-gamma
ˇ
1 have large variances, expressing your lack of knowledge
Simple Linear Regression
F
4105 about the regression coefficients. The priors correspond to an equal-tail 95 % credible intervals of approximately
. 2000; 2000/ for
ˇ
0 and
ˇ
1
. Priors of this type are often called vague or diffuse
priors. See the section “ Prior Distributions ” on page 142 for more information. Typically diffuse
prior distributions have little influence on the posterior distribution and are appropriate when stronger prior information about the parameters is not available.
A frequently used diffuse prior for the variance parameter
2 is the inverse-gamma distribution. With a shape parameter of credible interval of
3=10 and a scale parameter of
.1:7; 1e6/
, with the mode at
10=3 , this prior corresponds to an equal-tail
2:5641 for
95 %
2
. Alternatively, you can use any other positive prior, meaning that the density support is positive on this variance component. For example, you can use the gamma prior.
According to Bayes’ theorem, the likelihood function and prior distributions determine the posterior
(joint) distribution of ˇ
0
, ˇ
1
, and
2 as follows:
.ˇ
0
; ˇ
1
;
2 j
Weight ; Height /
/
.ˇ
0
/.ˇ
1
/.
2
/p.
Weight j
ˇ
0
; ˇ
1
;
2
; Height /
You do not need to know the form of the posterior distribution when you use PROC MCMC. PROC
MCMC automatically obtains samples from the desired posterior distribution, which is determined by the prior and likelihood you supply.
The following statements fit this linear regression model with diffuse prior information:
ods graphics on; proc mcmc data=class outpost=classout nmc=50000 thin=5 seed=246810; parms beta0 0 beta1 0; parms sigma2 1; prior beta0 beta1 ~ normal(mean = 0, var = 1e6); prior sigma2 ~ igamma(shape = 3/10, scale = 10/3); mu = beta0 + beta1*height; model weight ~ n(mu, var = sigma2); run; ods graphics off;
The ODS GRAPHICS ON statement invokes the ODS Graphics environment and displays the diagnostic plots, such as the trace and autocorrelation function plots of the posterior samples. For
more information about ODS, see Chapter 21, “ Statistical Graphics Using ODS .”
The PROC MCMC statement invokes the procedure and specifies the input data set data set classout class contains the posterior samples for all of the model parameters. The
. The output
option specifies the number of posterior simulation iterations. The
option controls the thinning of the Markov chain and specifies that one of every 5 samples is kept. Thinning is often used to reduce the correlations among posterior sample draws. In this example, 10,000 simulated values are saved in the classout data set. The
option specifies a seed for the random number generator, which guarantees the reproducibility of the random stream. For more information about Markov chain
sample size, burn-in, and thinning, see the section “ Burn-in, Thinning, and Markov Chain Samples ”
on page 154.
The
statements identify the three parameters in the model: beta0
, beta1
, and sigma2
. Each statement also forms a block of parameters, where the parameters are updated simultaneously in each
4106
F
Chapter 52: The MCMC Procedure iteration. In this example, beta0 and beta1 are sampled jointly, conditional on sigma2
; and sigma2 is sampled conditional on fixed values of beta0 and beta1
. In simple regression models such as this, you expect the parameters beta0 and beta1 to have high posterior correlations, and placing them both in the same block improves the mixing of the chain—that is, the efficiency that the posterior
statements also assign initial values to the parameters
given 0 as their initial values, and the scale parameter sigma2 starts at value 1. If you do not provide initial values, the procedure chooses starting values for every parameter.
The
statements specify prior distributions for the parameters. The parameters beta0 and beta1 both share the same prior—a normal prior with mean
0 and variance
1e6
. The parameter sigma2 has an inverse-gamma distribution with a shape parameter of 3/10 and a scale parameter of 10/3. For a
list of standard distributions that PROC MCMC supports, see the section “ Standard Distributions ”
on page 4155.
The mu assignment statement calculates the expected value of
Weight as a linear function of
Height
.
The
statement uses the shorthand notation,
n
, for the normal distribution to indicate that the response variable,
Weight
, is normally distributed with parameters mu and sigma2
. The functional argument MEAN= in the normal distribution is optional, but you have to indicate whether sigma2 is a variance (VAR=), a standard deviation (SD=), or a precision (PRECISION=) parameter. See
in the section “ MODEL Statement ” on page 4136 for distribution specifications.
The distribution parameters can contain expressions. For example, you can write the
statement as follows:
model weight ~ n(beta0 + beta1*height, var = sigma2);
Before you do any posterior inference, it is essential that you examine the convergence of the Markov
chain (see the section “ Assessing Markov Chain Convergence ” on page 155). You cannot make valid
inferences if the Markov chain has not converged. A very effective convergence diagnostic tool is the trace plot. Although PROC MCMC produces graphs at the end of the procedure output (see
), you should visually examine the convergence graph first.
The first table that PROC MCMC produces is the “Number of Observations” table, as shown in
. This table lists the number of observations read from the
data set and the number of non-missing observations used in the analysis.
Figure 52.1
Observation Information
Simple Linear Regression
The MCMC Procedure
Number of Observations Read
Number of Observations Used
19
19
The “Parameters” table, shown in
, lists the names of the parameters, the blocking
information (see the section “ Blocking of Parameters ” on page 4148), the sampling method used,
Simple Linear Regression
F
4107 distributions. You should to check this table to ensure that you have specified the parameters correctly, especially for complicated models.
Figure 52.2
Parameter Information
Parameters
Block
1
1
2
Parameter beta0 beta1 sigma2
Sampling
Method
N-Metropolis
N-Metropolis
N-Metropolis
Initial
Value Prior Distribution
0
0
1.0000
normal(mean = 0, var = 1e6) normal(mean = 0, var = 1e6) igamma(shape = 3/10, scale =
10/3)
The “Tuning History” table, shown in
, shows how the tuning stage progresses for the
multivariate random walk Metropolis algorithm used by PROC MCMC to generate samples from the posterior distribution. An important aspect of the algorithm is the calibration of the proposal distribution. The tuning of the Markov chain is broken into a number of phases. In each phase, PROC
MCMC generates trial samples and automatically modifies the proposal distribution as a result of the
PROC MCMC found an acceptable proposal distribution after 7 phases, and this distribution is used in both the burn-in and sampling stages of the simulation.
The “Burn-In History” table shows the burn-in phase, and the “Sampling History” table shows the main phase sampling.
Figure 52.3
Tuning, Burn-In and Sampling History
Tuning History
Phase
1
2
3
6
7
4
5
Block
2
1
2
1
2
1
2
1
1
2
1
2
1
2
Scale
2.3800
2.3800
1.0938
15.5148
0.8299
15.5148
1.1132
9.4767
1.4866
5.1914
2.2784
3.7859
2.8820
3.7859
Acceptance
Rate
0.0420
0.8860
0.2180
0.3720
0.4860
0.1260
0.4840
0.0880
0.5420
0.2000
0.4600
0.3900
0.3360
0.4020
4108
F
Chapter 52: The MCMC Procedure
Figure 52.3
continued
Block
Burn-In History
Scale
Acceptance
Rate
1
2
2.8820
3.7859
0.3400
0.4150
Block
Sampling History
Scale
Acceptance
Rate
1
2
2.8820
3.7859
0.3284
0.4008
For each posterior distribution, PROC MCMC also reports summary statistics (posterior means, standard deviations, and quantiles) and interval statistics (95% equal-tail and highest posterior density credible intervals), as shown in
. For more information about posterior statistics, see the
section “ Summary Statistics ” on page 169.
Figure 52.4
MCMC Summary and Interval Statistics
Parameter beta0 beta1 sigma2
N
10000
10000
10000
Simple Linear Regression
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-142.6
3.8917
136.8
33.9390
0.5427
51.7417
25%
Percentiles
50%
-164.5
3.5406
101.8
-142.4
3.8906
126.0
75%
-120.5
4.2402
159.9
Parameter beta0 beta1 sigma2
Alpha
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-209.3
2.8317
69.2208
-76.1692
4.9610
265.5
HPD Interval
-209.7
2.8280
58.2627
-77.1624
4.9468
233.8
By default, PROC MCMC also computes a number of convergence diagnostics to help you determine whether the chain has converged. These are the Monte Carlo standard errors, the autocorrelations at selected lags, the Geweke diagnostics, and the effective sample sizes. These statistics are shown in
. For details and interpretations of these diagnostics, see the section “ Assessing Markov
Chain Convergence ” on page 155.
Simple Linear Regression
F
4109
The “Monte Carlo Standard Errors” table indicates that the standard errors of the mean estimates for each of the parameters are relatively small, with respect to the posterior standard deviations.
The values in the “MCSE/SD” column (ratios of the standard errors and the standard deviations) are small, around 0.01. This means that only a fraction of the posterior variability is due to the simulation. The “Autocorrelations of the Posterior Samples” table shows that the autocorrelations among posterior samples reduce quickly and become almost nonexistent after lag 5. The “Geweke
Diagnostics” table indicates that no parameter failed the test, and the “Effective Sample Sizes” table reports the number of effective sample sizes of the Markov chain.
Figure 52.5
MCMC Convergence Diagnostics
Simple Linear Regression
The MCMC Procedure
Parameter
Monte Carlo Standard Errors
MCSE
Standard
Deviation beta0 beta1 sigma2
0.4576
0.00731
0.7151
33.9390
0.5427
51.7417
MCSE/SD
0.0135
0.0135
0.0138
Parameter beta0 beta1 sigma2
Posterior Autocorrelations
Lag 1 Lag 5 Lag 10
0.2986
0.2971
0.2966
-0.0008
0.0000
0.0062
0.0162
0.0135
0.0008
Geweke Diagnostics
Parameter z Pr > |z| beta0 beta1 sigma2
0.1105
-0.1701
-0.2175
0.9120
0.8649
0.8278
Lag 50
0.0193
0.0161
-0.0068
Parameter beta0 beta1 sigma2
Effective Sample Sizes
ESS
Autocorrelation
Time
5501.1
5514.8
5235.4
1.8178
1.8133
1.9101
Efficiency
0.5501
0.5515
0.5235
PROC MCMC produces a number of graphs, shown in
diagnostic checks. With the trace plots, there are two important aspects to examine. First, you want to check whether the mean of the Markov chain has stabilized and appears constant over the graph. Second, you want to check whether the chain has good mixing and is “dense,” in the sense
4110
F
Chapter 52: The MCMC Procedure that it quickly traverses the support of the distribution to explore both the tails and the mode areas efficiently. The plots show that the chains appear to have reached their stationary distributions.
Next, you should examine the autocorrelation plots, which indicate the degree of autocorrelation for each of the posterior samples. High correlations usually imply slow mixing. Finally, the kernel density plots estimate the posterior marginal distributions for each parameter.
Figure 52.6
Diagnostic Plots for
ˇ
0
,
ˇ
1 and
2
Figure 52.6
continued
Simple Linear Regression
F
4111
4112
F
Chapter 52: The MCMC Procedure
In regression models such as this, you expect the posterior estimates to be very similar to the maximum likelihood estimators with noninformative priors on the parameters, The REG procedure produces the following fitted model (code not shown):
Weight
D 143:0 C 3:9
Height
These are very similar to the means show in
. With PROC MCMC, you can carry
out informative analysis that uses specifications to indicate prior knowledge on the parameters.
Informative analysis is likely to produce different posterior estimates, which are the result of information from both the likelihood and the prior distributions. Incorporating additional information in the analysis is one major difference between the classical and Bayesian approaches to statistical inference.
One of the famous examples in the history of statistics is the Behrens-Fisher problem ( Fisher
1935 ). Consider the situation where there are two independent samples from two different normal
distributions: y
11
; y
12
; ; y
1n
1 y
21
; y
22
; ; y
2n
2 normal
.
1
;
2
1
/ normal
.
2
;
2
2
/
Note that hypothesis n
1
H
0
¤
W distribution under n
2
. When you do not want to assume that the variances are equal, testing the
1
H
0
D
2 is a difficult problem in the classical statistics framework, because the is not known. Within the Bayesian framework, this problem is straightforward because you can estimate the posterior distribution of
1 2 while taking into account the uncertainties in all of parameters by treating them as random variables.
Suppose that you have the following set of data:
title 'The Behrens-Fisher Problem'; data behrens; input y ind @@; datalines;
121 1 94 1 119 1 122 1 142 1 168 1 116 1
172 1 155 1 107 1 180 1 119 1 157 1 101 1
145 1 148 1 120 1 147 1 125 1 126 2 125 2
130 2 130 2 122 2 118 2 118 2 111 2 123 2
126 2 127 2 111 2 112 2 121 2
;
The response variable is y
, and the ind variable is the group indicator, which takes two values:
1 and
2
. There are 19 observations that belong to group
1 and 14 that belong to group
2
.
The Behrens-Fisher Problem
F
4113
The likelihood functions for the two samples are as follows: p.y
1i j p.y
2j j
1
;
2
1
/
2
;
2
2
/
D
D
.y
1i
I
.y
2j
I
1
;
2
1
/
2
;
2
2
/ for for i D 1; ; 19 j
D
1; ; 14
prior. The distribution is invariant under location transformations—that is,
D C c
. You can use this prior for the mean parameters in the model:
.
1
/
.
2
/
/
/
1
1
In addition,
Berger ( 1985 ) showed that a prior of the form
1= parameter, and it is invariant under scale transformations (that is
2 for the variance parameters in the model: is noninformative for the scale
D c
2
). You can use this prior
.
2
1
/
.
2
2
/
/
/
1=
2
1
1=
2
2
The log densities of the prior distributions on
2
1 and
2
2 are: log
..
log ..
2
1
2
2
//
//
D
D log
.
2
1
/ log .
2
2
/
The following statements generate posterior samples of means:
1 2
:
1
;
2
;
1
2
;
2
2
, and the difference in the
proc mcmc data=behrens outpost=postout seed=123 nmc=40000 thin=10 monitor=(_parms_ mudif) statistics(alpha=0.01)=(summary interval); ods select PostSummaries PostIntervals; parm mu1 0 mu2 0; parm sig21 1; parm sig22 1; prior mu: ~ general(0); prior sig21 ~ general(-log(sig21)); prior sig22 ~ general(-log(sig22)); mudif = mu1 - mu2; if ind = 1 then llike = lpdfnorm(y, mu1, sqrt(sig21)); else llike = lpdfnorm(y, mu2, sqrt(sig22)); model general(llike); run;
The PROC MCMC statement specifies an input data set ( behrens ), an output data set containing the posterior samples ( postout
), a random number seed, the simulation size, and the thinning rate. The
4114
F
Chapter 52: The MCMC Procedure
option specifies a list of symbols, which can be either parameters or functions of the parameters in the model, for which inference is to be done. The symbol
_parms_ is a shorthand for all model parameters—in this case, mu1
, mu2
, sig21
, and program as the difference between
1 and
2
.
sig22
. The symbol mudif is defined in the
The ODS SELECT statement displays the summary statistics and interval statistics tables while excluding all other output. For a complete list of ODS tables that PROC MCMC can produce, see
the sections “ Displayed Output
” on page 4195 and “ ODS Table Names ” on page 4200.
The
option calculates summary and interval statistics. The global suboption AL-
PHA=0.01 specifies 99 % equal-tail and highest posterior density (HPD) credible intervals for all parameters.
The
statements assign the parameters mu1 and mu2 to the same block, and sig21 and sig22 each to their own separate blocks. There are a total of three blocks. The
statements also assign an initial value to each parameter.
The
statements specify prior distributions for the parameters. Because the priors are all nonstandard (uniform on the real axis for
1 and
2 and
1=
2 for
2
1 and
2
2
), you must use the
function here. The argument in the
function is an expression for the log of the distribution, up to an additive constant. This distribution can have any functional form, as long as it is programmable using SAS functions and expressions. Note that the function specifies a distribution on the log scale, not the original scale. The log of the prior on mu1 and mu2 is
0
, and the log of the priors on sig21 and sig22 are
-log(sig21) and
-log(sig22) respectively. See the section
distribution.
The mudif assignment statement calculates the difference between mu1 and mu2
.
“ Using Density Functions in the Programming Statements ” on page 4167 for more details.
statement specifies that llike is the log likelihood for each observation in the model.
The
IF-ELSE statements enable different pending on their group indicator ind
.
y
’s to have different log-likelihood functions, de-
The function LPDFNORM is a PROC MCMC function that calculates the log density of a normal distribution.
See the section
The
displays the posterior summary and interval statistics.
Figure 52.7
Posterior Summary and Interval Statistics
Parameter mu1 mu2 sig21 sig22 mudif
N
4000
4000
4000
4000
4000
The Behrens-Fisher Problem
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
134.8
121.4
683.2
51.3975
13.3596
6.0065
1.9150
259.9
24.2881
6.3335
25%
Percentiles
50%
130.9
120.2
507.8
35.0212
9.1732
134.7
121.4
630.1
45.7449
13.4078
75%
138.7
122.7
792.3
61.2582
17.6332
The Behrens-Fisher Problem
F
4115
Figure 52.7
continued
Parameter mu1 mu2 sig21 sig22 mudif
Alpha
0.010
0.010
0.010
0.010
0.010
Posterior Intervals
Equal-Tail Interval
118.7
115.9
292.0
18.5883
-3.2537
150.6
126.6
1821.1
158.8
29.9987
HPD Interval
119.3
116.2
272.8
16.3730
-3.1915
151.0
126.7
1643.7
140.5
30.0558
The mean difference has a posterior mean value of 13:36 , and the lower endpoints of the 99 % credible intervals are negative. This suggests that the mean difference is positive with a high probability.
However, if you want to estimate the probability that
1 2
> 0
, you can do so as follows.
The following statements produce
proc format; value diffmt low-0 = 'mu1 - mu2 <= 0' 0<-high = 'mu1 - mu2 > 0'; run; proc freq data = postout; tables mudif /nocum; format mudif diffmt.; run;
The sample estimate of the posterior probability that
1 2
> 0 is 0.98. This example illustrates an advantage of Bayesian analysis. You are not limited to making inferences based on model parameters only. You can accurately quantify uncertainties with respect to any function of the parameters, and this allows for flexibility and easy interpretations in answering many scientific questions.
Figure 52.8
Estimated Probability of
1 2
> 0
.
The Behrens-Fisher Problem
The FREQ Procedure mudif Frequency Percent
--------------------------------------mu1 - mu2 <= 0 mu1 - mu2 > 0
77
3923
1.93
98.08
4116
F
Chapter 52: The MCMC Procedure
This example illustrates how you can fit a mixed-effects model in PROC MCMC. PROC MCMC
offers you the ability to model beyond the normal likelihood (see “ Example 52.5: Random-Effects
Models ” on page 4238), and you can model as many levels of random effects as are needed with this
procedure.
Consider a scenario in which data are collected in groups and you wish to model group-specific effects. You can use a mixed-effects model (sometimes also known as a random-effects model or a variance-components model): y ij
D ˇ
0
C ˇ
1 x ij
C i
C e ij
; e ij normal
.0;
2
/ where i D 1; 2; ; I is the group index and j group. In the regression model, the fixed effects
D
ˇ
0
1; 2; and
ˇ
1
; n i indexes the observations in the i th are the intercept and the coefficient for variable x ij
, respectively. The random effects i is the mean for the i th group, and e ij are the error term.
Consider the following SAS data set:
title 'Mixed-Effects Model'; data heights; input Family G$ Height @@; datalines;
1 F 67 1 F 66
2 F 63
3 M 64
2 F 67
4 F 67
1 F 64
2 M 69
4 F 66
1 M 71
2 M 68
4 M 67
;
1 M 72
2 M 70
4 M 67
2 F 63
3 F 63
4 M 69 data input; set heights; if g eq 'F' then gender = 1; else gender = 0; drop g; run;
The response variable
Height measures the heights (in inches) of 18 individuals. The individuals are classified according to
Family and
Gender
.
Height is assumed to be normally distributed: y ij normal .
ij
;
2
/; ij
D ˇ
0
C ˇ
1 x ij
C which corresponds to a normal likelihood as follows: p.y
ij j ij
;
2
/ D .
ij
; var
D
2
/ i
The priors on the parameters ˇ
0
, ˇ
1
, i are assumed to be normal as well:
.ˇ
0
/
.ˇ
1
/
.
i
/
D
D
D
.0; var
D 1e5/
.0; var
.0; var
D 1e5/
D
2
/
Mixed-Effects Model
F
4117
Priors on the variance terms,
2 and
2
, are inverse-gamma:
.
2
/
.
2
/
D
D f i
.
shape
D 0:001; scale
D 1000/ f i
.
shape
D 0:001; scale
D 1000/ where f i denotes the density function of an inverse-gamma distribution.
The following statements fit a linear random-effects model to the data and produce the output shown in
and
ods graphics on; proc mcmc data=input outpost=postout thin=10 nmc=50000 seed=7893 monitor=(b0 b1); ods select PostSummaries PostIntervals tadpanel; array gamma[4]; parms b0 0 b1 0 gamma: 0; parms s2 1 ; parms s2g 1; prior b: ~ normal(0, var = 10000); prior gamma: ~ normal(0, var = s2g); prior s2: ~ igamma(0.001, scale = 1000); mu = b0 + b1 * gender + gamma[family]; model height ~ normal(mu, var = s2); run; ods graphics off;
The statements are very similar to those shown in the previous two examples. The ODS GRAPHICS
ON statement requests ODS Graphics. The PROC MCMC statement specifies the input and output data sets, the simulation size, the thinning rate, and a random number seed. The
option indicates that the model parameters b0 and b1 are the quantities of interest. The ODS SELECT statement displays the summary statistics table, the interval statistics table, and the diagnostics plots.
The
statement defines a one-dimensional array, gamma
, with 4 elements. You can refer to the array elements with variable names ( gamma1 to gamma4 by default) or with subscripts, such as gamma[2]
. To indicate subscripts, you must use either brackets
Πor braces f g
, but not parentheses
. /
. Note that this is different from the way subscripts are indicated in the DATA step. See the section
“ ARRAY Statement ” on page 4132 for more information.
The
statements specify priors for all the parameters. The notation symbols that start with the letter ‘b’. In this example, it includes b0 and b1 b: is a shorthand for all
. Similarly, gamma: stands for all four gamma parameters, and s2: stands for both s2 and s2g
. This shorthand notation can save you some typing, and it keeps your statements tidy.
The mu assignment statement calculates the expected value of height in the random-effects model.
The symbol family is a data set variable that indexes family. Here gamma[family] is the random effect for the value of family
.
Finally, the
statement specifies the likelihood function for height
.
4118
F
Chapter 52: The MCMC Procedure
The posterior summary and interval statistics for b0 and b1 are shown in
Figure 52.9
Posterior Summary and Interval Statistics
Parameter b0 b1
N
5000
5000
Mixed-Effects Model
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
66.2685
-3.3492
19.1176
6.3886
25%
Percentiles
50%
56.0024
-7.4268
66.7260
-3.2799
75%
77.2356
0.6078
Parameter b0 b1
Alpha
0.050
0.050
Posterior Intervals
Equal-Tail Interval
26.2226
-16.2018
103.3
9.6267
HPD Interval
27.1749
-17.0757
103.6
8.5265
Trace plots, autocorrelation plots, and posterior density plots for b1 and logpost are shown in
b1 looks good. The convergence plots for the other parameters also look reasonable, and are not shown here.
Figure 52.10
Plots for b
1 and Log of the Posterior Density
Figure 52.10
continued
Syntax: MCMC Procedure
F
4119
From the interval statistics table, you see that both the equal-tail and HPD intervals for ˇ
0 are positive, strongly indicating the positive effect of the parameter. On the other hand, both intervals for
ˇ
1 cover the value zero, indicating that this model.
gender does not have a strong impact on predicting height in
The following statements are available in PROC MCMC. Items within < > are optional.
< options >
;
arrayname <{ dimensions }>
;
;
;
variables
;
variable distribution
;
parameter < = > number < /options >
;
< ’label’ >
OUTPRED=
SAS-data-set < options >
;
;
parameter distribution
;
subroutine-name ( subroutine-argument-list)
;
4120
F
Chapter 52: The MCMC Procedure
The
statements declare parameters in the model and assign optional starting values for the Markov chain. The
statements specify the prior distributions of the parameters. The
statements specify the log-likelihood functions for the response variables.
These statements form the basis of every Bayesian model.
In addition, you can use the
statement to define constant or parameter arrays, the
and similar statements to save unnecessary evaluation and reduce simulation time, the
statement to generate samples from the posterior predictive distribution, the
to specify more complicated models that you wish to fit, and finally the
statements to define your own Gibbs samplers to sample any parameters in the model.
The following sections provide a description of each of these statements.
PROC MCMC
options
;
This statement invokes PROC MCMC.
A number of options are available in the PROC MCMC statement; the following table categorizes them according to function.
Table 52.1
PROC MCMC Statement Options
Option Description
Basic options
Debugging output
names the input data set names the output data set for posterior samples of parameters displays model program and variables displays compiled model program displays detailed model execution messages
Frequently used MCMC options
specifies the maximum number of tuning loops
specifies the minimum number of tuning loops specifies the number of burn-in iterations specifies the number of MCMC iterations, excluding the burn-in iterations
specifies the number of tuning iterations controls options for constructing the initial proposal covariance matrix specifies the random seed for simulation specifies the thinning rate
Less frequently used MCMC options
specifies the acceptance rate tolerance controls sampling discrete parameters controls generating initial values
PROC MCMC Statement
F
4121
Table 52.1
(continued)
Option Description
Other Options
specifies the proposal distribution specifies the initial scale applied to the proposal distribution specifies the target acceptance rate for random walk sampler specifies the target acceptance rate for independence sampler specifies the weight used in covariance updating
Summary, diagnostics, and plotting options
specifies the number of autocorrelation lags used to compute effective
sample sizes and Monte Carlo errors controls the convergence diagnostics computes deviance information criterion (DIC)
outputs analysis for a list of symbols of interest controls plotting controls posterior statistics specifies the machine numerical limit for infinity specifies joint log-likelihood function indicates how missing values are handled.
controls the frequency of report for expected run time specifies the singularity tolerance
These options are described in alphabetical order.
ACCEPTTOL=
n specifies a tolerance for acceptance probabilities. By default, ACCEPTTOL=0.075.
AUTOCORLAG=
n
ACLAG=
n specifies the maximum number of autocorrelation lags used in computing the effective sample
where MCsample is the Markov chain sample size kept after thinning—that is, MCsample
D h
NMC
NTHIN i
. If AUTOCORLAG= is set too low, you might observe significant lags, and the effective sample size cannot be calculated accurately. A WARNING message appears, and you can either increase AUTOCORLAG= or
DISCRETE=
keyword specifies the proposal distribution used in sampling discrete parameters. The default is
DISCRETE=BINNING.
The keyword values are as follows:
BINNING
uses continuous proposal distributions for all discrete parameter blocks. The proposed
4122
F
Chapter 52: The MCMC Procedure sample is then discretized (binned) before further calculations. This sampling method approximates the correlation structure among the discrete parameters in the block and could improve mixing in some cases.
GEO
uses independent symmetric geometric proposal distributions for all discrete parameter blocks. This proposal does not take parameter correlations into account. However, it can work better than the BINNING option in cases where the range of the parameters is relatively small and a normal approximation can perform poorly.
DIAGNOSTICS=NONE |
(keyword-list)
DIAG=NONE |
(keyword-list) specifies options for MCMC convergence diagnostics. By default, PROC MCMC computes the Geweke test, sample autocorrelations, effective sample sizes, and Monte Carlo errors. The
Raftery-Lewis and Heidelberger-Welch tests are also available. See the section “ Assessing
Markov Chain Convergence ” on page 155 for more details on convergence diagnostics. You
can request all of the diagnostic tests by specifying DIAGNOSTICS=ALL. You can suppress all the tests by specifying DIAGNOSTICS=NONE.
The following options are available.
ALL
computes all diagnostic tests and statistics. You can combine the option ALL with any other specific tests to modify test options. For example DIAGNOSTICS=(ALL AUTO-
CORR(LAGS=(1 5 35))) computes all tests with default settings and autocorrelations at lags 1, 5, and 35.
AUTOCORR <
(autocorr-options)
>
computes default autocorrelations at lags 1, 5, 10, and 50 for each variable. You can choose other lags by using the following autocorr-options :
LAGS | AC=
numeric-list specifies autocorrelation lags. The numeric-list must take positive integer values.
ESS
computes the effective sample sizes ( Kass et al.
( 1998 )) of the posterior samples of each
parameter. It also computes the correlation time and the efficiency of the chain for each parameter. Small values of ESS might indicate a lack of convergence. See the section
“ Effective Sample Size ” on page 168 for more details.
GEWEKE <
(Geweke-options)
>
computes the Geweke spectral density diagnostics; this is a two-sample t
-test between the first f
1 portion and the last f
2
portion of the chain. See the section “ Geweke
Diagnostics ” on page 162 for more details. The default is FRAC1=0.1 and FRAC2=0.5,
but you can choose other fractions by using the following Geweke-options :
FRAC1 | F1=
value specifies the beginning FRAC1 proportion of the Markov chain. By default,
FRAC1=0.1.
PROC MCMC Statement
F
4123
FRAC2 | F2=
value specifies the end FRAC2 proportion of the Markov chain. By default, FRAC2=0.5.
HEIDELBERGER | HEIDEL <
(Heidel-options)
>
computes the Heidelberger and Welch diagnostic (which consists of a stationarity test and a halfwidth test) for each variable. The stationary diagnostic test tests the null hypothesis that the posterior samples are generated from a stationary process. If the stationarity test
is passed, a halfwidth test is then carried out. See the section “ Heidelberger and Welch
Diagnostics ” on page 164 for more details.
These diagnostics are not performed by default. You can specify the DIAGNOS-
TICS=HEIDELBERGER option to request these diagnostics, and you can also specify suboptions, such as DIAGNOSTICS=HEIDELBERGER(EPS=0.05), as follows:
SALPHA=
value specifies the
˛
PHA=0.05.
level
.0 < ˛ < 1/ for the stationarity test. By default, SAL-
HALPHA=
value specifies the ˛
PHA=0.05.
level .0 < ˛ < 1/ for the halfwidth test. By default, HAL-
EPS=
value specifies a small positive number such that if the halfwidth is less than times the sample mean of the retaining iterates, the halfwidth test is passed. By default,
EPS=0.1.
MCSE
MCERROR
computes the Monte Carlo standard error for the posterior samples of each parameter.
NONE
suppresses all of the diagnostic tests and statistics. This is not recommended.
RAFTERY | RL < (
Raftery-options
) >
computes the Raftery and Lewis diagnostics, which evaluate the accuracy of the estimated quantile (
O
Q for a given Q
2 .0; 1/
) of a chain.
O
Q can achieve any degree of accuracy when the chain is allowed to run for a long time. The algorithm stops when the estimated probability P
O
Q
D Pr .
O
Q
/ reaches within ˙ R of the value Q with probability S; that is, Pr
.
Q R
O
Q
Q
C
R
/ D
S. See the section “ Raftery and
Lewis Diagnostics ” on page 165 for more details. The
Raftery-options enable you to specify Q, R, S, and a precision level for a stationary test.
These diagnostics are not performed by default. You can specify the DIAGNOS-
TICS=RAFERTY option to request these diagnostics, and you can also specify suboptions, such as DIAGNOSTICS=RAFERTY(QUANTILE=0.05), as follows:
4124
F
Chapter 52: The MCMC Procedure
QUANTILE | Q=
value specifies the order (a value between 0 and 1) of the quantile of interest. By default,
QUANTILE=0.025.
ACCURACY | R=
value specifies a small positive number as the margin of error for measuring the accuracy of estimation of the quantile. By default, ACCURACY=0.005.
PROB | S=
value specifies the probability of attaining the accuracy of the estimation of the quantile.
By default, PROB=0.95.
EPS=
value specifies the tolerance level (a small positive number) for the stationary test. By default, EPS=0.001.
DIC
computes the Deviance Information Criterion (DIC). DIC is calculated using the posterior
mean estimates of the parameters. See the section “ Deviance Information Criterion (DIC) ” on
page 171 for more details.
DATA=
SAS-data-set specifies the input data set. Observations in this data set are used to compute the log-likelihood function that you specify with PROC MCMC statements.
INF=
value specifies the numerical definition of infinity in the procedure. The default is INF= 1 E 15 . For example, PROC MCMC considers
1
E
16 to be outside of the support of the normal distribution and assigns a missing value to the log density evaluation. You can select a larger value with the INF= option. The minimum value allowed is 1 E 10 .
INIT=
(keyword-list) specifies options for generating the initial values for the parameters. These options apply only
or
is used, you must supply explicit initial values for the parameters. By default,
INIT=MODE. The following keywords are used:
MODE
uses the mode of the prior density as the initial value of the parameter, if you did not provide one. If the mode does not exist or if it is on the boundary of the support of the density, the mean value is used. If the mean is outside of the support or on the boundary, which can happen if the prior distribution is truncated, a random number drawn from the prior is used as the initial value.
PROC MCMC Statement
F
4125
PINIT
tabulates parameter values after the tuning phase. This option also tabulates the tuned proposal parameters used by the Metropolis algorithm. These proposal parameters include covariance matrices for continuous parameters and probability vectors for discrete parameters for each block. By default, PROC MCMC does not display the initial values or the tuned proposal parameters after the tuning phase.
RANDOM
generates a random number from the prior density and uses it as the initial value of the parameter, if you did not provide one.
REINIT
resets the parameters, after the tuning phase, with the initial values that you provided explicitly or that were assigned by the procedure. By default, PROC MCMC does not reset the parameters because the tuning phase usually moves the Markov chains to a more favorable place in the posterior distribution.
LIST
displays the model program and variable lists. The LIST option is a debugging feature and is not normally needed.
LISTCODE
displays the compiled program code. The LISTCODE option is a debugging feature and is not normally needed.
JOINTMODEL
JOINTLLIKE
specifies how the likelihood function is calculated. By default, PROC MCMC assumes that the observations in the data set are independent so that the joint log-likelihood function is the sum of the individual log-likelihood functions for the observations, where the individual loglikelihood function is specified in the
statement. When your data are not independent, you can specify the JOINTMODEL option to modify the way that PROC MCMC computes the joint log-likelihood function. In this situation, PROC MCMC no longer steps through the input data set to sum the individual log likelihood.
To use this option correctly, you need to do the following two things: create ARRAY symbols to store all data set variables that are used in the program. This can be accomplished with the
and
statements.
program the joint log-likelihood function by using these ARRAY symbols only. The
statement specifies the joint log-likelihood function for the entire data set.
Typically, you use the function
in the
statement.
See the sections “ BEGINCNST/ENDCNST Statement
” on page 4133 and “ Modeling Joint
Likelihood ” on page 4181 for details.
MAXTUNE=
n specifies an upper limit for the number of proposal tuning loops. By default, MAXTUNE=24.
See the section “ Covariance Tuning ” on page 4151 for more details.
4126
F
Chapter 52: The MCMC Procedure
MINTUNE=
n specifies a lower limit for the number of proposal tuning loops. By default, MINTUNE=2. See
the section “ Covariance Tuning ” on page 4151 for more details.
MISSING=
keyword
MISS=
keyword
specifies how missing values are handled (see the section “ Handling of Missing Data ” on
page 4190 for more details). The default is MISSING=COMPLETECASE.
ALLCASE | AC
gives you the option to model the missing values in an all-case analysis. You can use any techniques that you see fit, for example, fully Bayesian or multiple imputation.
COMPLETECASE | CC
assumes a complete case analysis, so all observations with missing variable values are discarded prior to the simulation.
MONITOR= (
symbol-list
)
outputs analysis for selected symbols of interest in the program. The symbols can be any of the following: model parameters (symbols in the
statement), secondary parameters
(assigned using the operator “=”), the log of the posterior density (LOGPOST), the log of the prior density (LOGPRIOR), the log of the hyperprior density (LOGHYPER) if the
statement is used, or the log of the likelihood function (LOGLIKE). You can use the keyword
_PARMS_ as a shorthand for all of the model parameters. PROC MCMC performs only posterior analyses (such as plotting, diagnostics, and summaries) on the symbols selected with the MONITOR= option. You can also choose to monitor an entire array by specifying the name of the array. By default MONITOR=_PARMS_.
Posterior samples of any secondary parameters listed in the MONITOR= option are saved in the
data set. Posterior samples of model parameters are always saved to the
data set, regardless of whether they appear in the MONITOR= option.
NBI=
n specifies the number of burn-in iterations to perform before beginning to save parameter
estimate chains. By default, NBI=1000. See the section “ Burn-in, Thinning, and Markov
Chain Samples ” on page 154 for more details.
NMC=
n specifies the number of iterations in the main simulation loop. This is the MCMC sample size if
THIN=1 . By default, NMC=1000.
NTU=
n specifies the number of iterations to use in each proposal tuning phase. By default, NTU=500.
OUTPOST=
SAS-data-set specifies an output data set that contains the posterior samples of all model parameters, the iteration numbers (variable name ITERATION), the log of the posterior density (LOGPOST), the log of the prior density (LOGPRIOR), the log of the hyperprior density (LOGHYPER), if the
statement is used, and the log likelihood (LOGLIKE). Any secondary parameters
PROC MCMC Statement
F
4127
(assigned using the operator “=”) listed in the
option are saved to this data set.
By default, no OUTPOST= data set is created.
PLOTS
< (global-plot-options) >
=
(plot-request < . . . plot-request > )
PLOT
< (global-plot-options) >
=
(plot-request < . . . plot-request > ) controls the display of diagnostic plots. Three types of plots can be requested: trace plots, autocorrelation function plots, and kernel density plots. By default, the plots are displayed in panels unless the global plot option UNPACK is specified. Also when more than one type of plot is specified, the plots are grouped by parameter unless the global plot option
GROUPBY=TYPE is specified. When you specify only one plot request, you can omit the parentheses around the plot-request, as shown in the following example:
plots=none plots(unpack)=trace plots=(trace density)
You must enable ODS Graphics before requesting plots—for example, like this:
ods graphics on; proc mcmc;
...; run; ods graphics off;
If you have enabled ODS Graphics but do not specify the PLOTS= option, then PROC MCMC produces, for each parameter, a panel that contains the trace plot, the autocorrelation function plot, and the density plot. This is equivalent to specifying PLOTS=(TRACE AUTOCORR
DENSITY).
The global-plot-options include the following:
FRINGE
adds a fringe plot to the horizontal axis of the density plot.
GROUPBY|GROUP=PARAMETER | TYPE
specifies how the plots are grouped when there is more than one type of plot.
GROUPBY=PARAMETER is the default. The choices are as follows:
TYPE
specifies that the plots are grouped by type.
PARAMETER
specifies that the plots are grouped by parameter.
LAGS=
n specifies the number of autocorrelation lags used in plotting the ACF graph. By default,
LAGS=50.
4128
F
Chapter 52: The MCMC Procedure
SMOOTH
smoothes the trace plot with a fitted penalized B-spline curve ( Eilers and Marx 1996 ).
UNPACKPANEL
UNPACK
specifies that all paneled plots are to be unpacked, so that each plot in a panel is displayed separately.
The plot-requests are as follows:
ALL
requests all types of plots. PLOTS=ALL is equivalent to specifying PLOTS=(TRACE
AUTOCORR DENSITY).
AUTOCORR | ACF
displays the autocorrelation function plots for the parameters.
DENSITY | D | KERNEL | K
displays the kernel density plots for the parameters.
NONE
suppresses the display of all plots.
TRACE | T
displays the trace plots for the parameters.
Consider a model with four parameters, X1–X4. Displays for various specifications are depicted as follows.
PLOTS=(TRACE AUTOCORR) displays the trace and autocorrelation plots for each parameter side by side with two parameters per panel:
Display 1 Trace(X1)
Trace(X2)
Autocorr(X1)
Autocorr(X2)
Display 2 Trace(X3)
Trace(X4)
Autocorr(X3)
Autocorr(X4)
PLOTS(GROUPBY=TYPE)=(TRACE AUTOCORR) displays all the paneled trace plots, followed by panels of autocorrelation plots:
Display 1 Trace(X1)
Trace(X2)
Display 2 Trace(X3)
Trace(X4)
Display 3 Autocorr(X1)
Autocorr(X3)
Autocorr(X2)
Autocorr(X4)
PROC MCMC Statement
F
4129
PLOTS(UNPACK)=(TRACE AUTOCORR) displays a separate trace plot and a separate correlation plot, parameter by parameter:
Display 1 Trace(X1)
Display 2 Autocorr(X1)
Display 3 Trace(X2)
Display 4 Autocorr(X2)
Display 5 Trace(X3)
Display 6 Autocorr(X3)
Display 7 Trace(X4)
Display 8 Autocorr(X4)
PLOTS(UNPACK GROUPBY=TYPE)=(TRACE AUTOCORR) displays all the separate trace plots followed by the separate autocorrelation plots:
Display 1 Trace(X1)
Display 2 Trace(X2)
Display 3 Trace(X3)
Display 4 Trace(X4)
Display 5 Autocorr(X1)
Display 6 Autocorr(X2)
Display 7 Autocorr(X3)
Display 8 Autocorr(X4)
PROPCOV=
value specifies the method used in constructing the initial covariance matrix for the Metropolis-
Hastings algorithm. The QUANEW and NMSIMP methods find numerically approximated covariance matrices at the optimum of the posterior density function with respect to all continuous parameters. The optimization does not apply to discrete parameters. The tuning phase starts at the optimized values; in some problems, this can greatly increase convergence performance. If the approximated covariance matrix is not positive definite, then an identity matrix is used instead. Valid values are as follows:
4130
F
Chapter 52: The MCMC Procedure
IND
uses the identity covariance matrix. This is the default. See the section “ Tuning the
Proposal Distribution ” on page 4150.
CONGRA< (
optimize-options
) >
performs a conjugate-gradient optimization.
DBLDOG< (
optimize-options
) >
performs a double-dogleg optimization.
QUANEW< (
optimize-options
) >
performs a quasi-Newton optimization.
NMSIMP | SIMPLEX< (
optimize-options
) >
performs a Nelder-Mead simplex optimization.
The optimize-options are as follows:
ITPRINT
prints optimization iteration steps and results.
PROPDIST=
value
specifies a proposal distribution for the Metropolis algorithm. See the section “ Metropolis and
Metropolis-Hastings Algorithms ” on page 150. You can also use
statement option
(see the section “ PARMS Statement ” on page 4139) to change the proposal distribution for a
particular block of parameters. Valid values are as follows:
NORMAL
N
specifies a normal distribution as the proposal distribution. This is the default.
T< (
df
) >
specifies a t -distribution with the degrees of freedom df . By default, df =3. If df the normal distribution is used since the two distributions are almost identical.
> 100
,
SCALE=
value controls the initial multiplicative scale to the covariance matrix of the proposal distribution.
By default, SCALE=2.38. See the section “ Scale Tuning ” on page 4151 for more details.
SEED=
n specifies the random number seed. By default, SEED=0, and PROC MCMC gets a random number seed from the clock.
SIMREPORT=
n controls the number of times that PROC MCMC reports the expected run time of the simulation.
This can be useful for monitoring the progress of CPU-intensive programs. For example, with SIMREPORT=2, PROC MCMC reports the simulation progress twice. By default,
SIMREPORT=0, and there is no reporting. The expected run times are displayed in the log file.
PROC MCMC Statement
F
4131
SINGDEN=
value defines the singularity criterion in the procedure. By default, SINGDEN=1E-11. The value indicates the exclusion of an endpoint in an interval. The mathematical notation “
.0
” is equivalent to “ Œ value ” in PROC MCMC—that is, procedure. The maximum SINGDEN allowed is
1
E x < 0
6
.
is treated as x value in the
STATISTICS
< (global-stats-options) >
= NONE | ALL |
stats-request
STATS
< (global-stats-options) >
= NONE | ALL |
stats-request specifies options for posterior statistics. By default, PROC MCMC computes the posterior mean, standard deviation, quantiles, and two
95
% credible intervals: equal-tail and highest posterior density (HPD). Other available statistics include the posterior correlation and covari-
ance. See the section “ Summary Statistics ” on page 169 for more details. You can request all
of the posterior statistics by specifying STATS=ALL. You can suppress all the calculations by specifying STATS=NONE.
The global-stats-options includes the following:
ALPHA=
numeric-list specifies the
0 and
˛ level for the equal-tail and HPD intervals. The value
0:5
. By default, ALPHA=0.05.
˛ must be between
PERCENTAGE | PERCENT=
numeric-list calculates the posterior percentages. The numeric-list
100
. By default, PERCENTAGE=(25 50 75).
contains values between 0 and
The stats-requests include the following:
ALL
computes all posterior statistics. You can combine the option ALL with any other options. For example STATS(ALPHA=(0.02 0.05 0.1))=ALL computes all statistics with the default settings and intervals at ˛ levels of 0.02, 0.05, and 0.1.
CORR
computes the posterior correlation matrix.
COV
computes the posterior covariance matrix.
SUMMARY
SUM
computes the posterior means, standard deviations, and percentile points for each variable. By default, the 25th, 50th, and 75th percentile points are produced, but you can use the global PERCENT= option to request specific percentile points.
INTERVAL
INT
computes the
100.1
˛/
% equal-tail and HPD credible intervals for each variable. See
the sections “ Equal-Tail Credible Interval ” on page 170 and “ Highest Posterior Density
(HPD) Interval ” on page 170 for details. By default, ALPHA=0.05, but you can use the
global ALPHA= option to request other intervals of any probabilities.
4132
F
Chapter 52: The MCMC Procedure
NONE
suppresses all of the statistics.
TARGACCEPT=
value specifies the target acceptance rate for the random walk based Metropolis algorithm. See the
section “ Metropolis and Metropolis-Hastings Algorithms ” on page 150. The numeric
value must be between 0:01 and 0:99 . By default, TARGACCEPT=0.45 for models with 1 parameter;
TARGACCEPT=0.35 for models with 2, 3, or 4 parameters; and TARGACCEPT=0.234 for
models with more than 4 parameters ( Roberts, Gelman, and Gilks 1997 ; Roberts and Rosenthal
TARGACCEPTI=
value specifies the target acceptance rate for the independence sampler algorithm. The independence
sampler is used for blocks of binary parameters. See the section “ Independence Sampler ”
on page 153 for more details. The numeric
TARGACCEPTI=0.6.
value must be between
0 and
1
. By default,
THIN=
n
NTHIN=
n controls the thinning rate of the simulation. PROC MCMC keeps every n th simulation sample and discards the rest. All of the posterior statistics and diagnostics are calculated using the
thinned samples. By default, THIN=1. See the section “ Burn-in, Thinning, and Markov Chain
Samples ” on page 154 for more details.
TRACE
displays the result of each operation in each statement in the model program as it is executed.
This debugging option is very rarely needed, and it produces voluminous output. If you use this option, also use small NMC=, NBI=, MAXTUNE=, and NTU= numbers.
TUNEWT=
value specifies the multiplicative weight used in updating the covariance matrix of the proposal distribution. The numeric value must be between 0 and 1. By default, TUNEWT=0.75. See
the section “ Covariance Tuning ” on page 4151 for more details.
ARRAY
arrayname <{ dimensions }> <$> <variables and constants>
;
The ARRAY statement associates a name (of no more than eight characters) with a list of variables and constants. The ARRAY statement is similar to, but not the same as, the ARRAY statement in the DATA step, and it is the same as the ARRAY statements in the NLIN, NLP, NLMIXED, and
MODEL procedures. The array name is used with subscripts in the program to refer to the array elements, as illustrated in the following statements:
array r[8] r1-r8; do i = 1 to 8;
BEGINCNST/ENDCNST Statement
F
4133
r[i] = 0; end;
The ARRAY statement does not support all the features of the ARRAY statement in the DATA step. Implicit indexing of variables cannot be used; all array references must have explicit subscript expressions. Only exact array dimensions are allowed; lower-bound specifications are not supported.
A maximum of six dimensions is allowed.
Both variables and constants can be array elements. Constant array elements cannot have values assigned to them while variables can. Both the dimension specification and the list of elements are optional, but at least one must be specified. When the list of elements is not specified or fewer elements than the size of the array are listed, array variables are created by appending element numbers to the array name to complete the element list. You can index array elements by enclosing a subscript in braces
.
f g / for function calls only.
or brackets
.Π/
, but not in parentheses
.. //
. The parentheses are reserved
For example, the following statement names an array day
:
array day[365];
By default, the variables names are day1 to day365
. However, since
day
is a SAS function, any subscript that uses parentheses gives you the wrong results. The expression
day(4)
returns the value
5 and does not reference the array element day4
.
BEGINCNST ;
ENDCNST ;
The BEGINCNST and ENDCNST statements define a block within which PROC MCMC processes the programming statements only during the setup stage of the simulation. You can use the BEGINC-
NST and ENDCNST statements to define constants or import data set variables into arrays. Storing data in arrays enables you to work with data that are not identically distributed (see the section
“ Modeling Joint Likelihood ” on page 4181) or to implement your own Markov chain sampler (see
the section “ UDS Statement ” on page 4143). You can also use the BEGINCNST and ENDCNST
page 4154).
Assign Constants
Whenever you have programming statements that calculate constants that do not need to be evaluated multiple times throughout the simulation, you should put them within the BEGINCNST and
ENDCNST statements. Using these statements can reduce redundant processing. For example, you can assign a constant to a symbol or fill in an array with numbers:
4134
F
Chapter 52: The MCMC Procedure
array cnst[17]; begincnst; offset = 17; do i = 1 to 17; cnst[i] = i * i; end; endcnst;
The MCMC procedure evaluates the programming statements with the BEGINCNST/ENDCNST block once and ignores them in the rest of the simulation.
READ_ARRAY Function
Sometimes you might need to store variables, either from the current input data set or from a different data set, in arrays and use these arrays to specify your model. The READ_ARRAY function is a convenient for that purpose.
The following two forms of the READ_ARRAY function are available:
rc = READ_ARRAY
(data_set, array)
; rc = READ_ARRAY
(data_set, array < ,"col_name_1" > < , "col_name_2" > < , ...
> )
;
where rc returns 0 if the function is able to successfully read the data set.
data_set specifies the name of the data set from which the array data is read. The value specified for data_set must be a character literal or a variable that contains the member name
(libname.memname) of the data set to be read from.
array specifies the PROC MCMC array variable into which the data is read. The value specified for array must be a local temporary array variable because the function might need to grow or shrink its size to accommodate the size of the data set.
col_name specifies optional names for the specific columns of the data set that are read. If specified, col_name must be a literal string enclosed in quotation marks. In addition, col_name cannot be a PROC MCMC variable. If column names are not specified, PROC MCMC reads all of the columns in the data set.
When SAS translates between an array and a data set, the array is indexed as [row,column].
The READ_ARRAY function attempts to dynamically resize the array to match the dimensions of the input data set. Therefore, the array must be dynamic; that is, the array must be declared with the
/NOSYMBOLS option.
For examples that use the READ_ARRAY function, see “ Modeling Joint Likelihood ” on page 4181,
” on page 4273, and “ Example 52.11: Implement a New Sampling
BEGINNODATA/ENDNODATA Statements
F
4135
BEGINNODATA ;
ENDNODATA ;
BEGINPRIOR ;
ENDPRIOR ;
The BEGINNODATA and ENDNODATA statements define a block within which PROC MCMC processes the programming statements without stepping through the entire data set. The programming statements are executed only twice: at the first and the last observation of the data set. The
BEGINNODATA and ENDNODATA statements are best used to reduce unnecessary observationlevel computations. Any computations that are identical to every observation, such as transformation of parameters, should be enclosed in these statements.
The BEGINPRIOR and ENDPRIOR statements are aliases for the BEGINNODATA and ENDNO-
DATA statements, respectively. You can enclose PRIOR statements in the BEGINNODATA and
ENDNODATA statements.
BY
variables
;
You can specify a BY statement with PROC PROC MCMC to obtain separate analyses on observations in groups that are defined by the BY variables. When a BY statement appears, the procedure expects the input data set to be sorted in order of the BY variables. If you specify more than one BY statement, only the last one specified is used.
If your input data set is not sorted in ascending order, use one of the following alternatives:
Sort the data by using the SORT procedure with a similar BY statement.
Specify the NOTSORTED or DESCENDING option in the BY statement for the PROC
MCMC procedure. The NOTSORTED option does not mean that the data are unsorted but rather that the data are arranged in groups (according to values of the BY variables) and that these groups are not necessarily in alphabetical or increasing numeric order.
Create an index on the BY variables by using the DATASETS procedure (in Base SAS software).
For more information about BY-group processing, see the discussion in SAS Language Reference:
Concepts . For more information about the DATASETS procedure, see the discussion in the Base
SAS Procedures Guide .
4136
F
Chapter 52: The MCMC Procedure
MODEL
dependent-variable-list distribution
;
The MODEL statement is used to specify the conditional distribution of the data given the parameters
(the likelihood function). You must specify a single dependent variable or a list of dependent variables, a tilde ( ), and then a distribution with its arguments. The dependent variables can be variables from the input data set or functions of the symbols in the program. The dependent variables must be specified unless the functions
or
are used (see the section
“ Specifying a New Distribution ” on page 4166 for more details). Multiple MODEL statements are
allowed for defining models with multiple independent components. The log likelihood value is the sum of the log likelihood values from each MODEL statement.
PROC MCMC is a programming language that is similar to the DATA step, and the order of statement evaluation is important. For example, the MODEL statement must come after any SAS programming statements that define or modify arguments used in the construction of the log likelihood. In PROC
MCMC, a symbol is allowed to be defined multiple times and used at different places. Using an expression out of order produces erroneous results that can also be hard to detect.
Standard distributions that the MODEL statement supports are listed in the
(see the
section “ Standard Distributions ” on page 4155 for density specification). These distributions can
also be used in the PRIOR and HYPERPRIOR statements. PROC MCMC allows some distributions to be parameterized in multiple ways. For example, you can specify a normal distribution with variance (VAR=), standard deviation (SD=), or precision (PRECISION=) parameter. For distributions that have different parameterizations, you must specify an option to clearly name the ambiguous parameter. In the normal distribution, for example, you must indicate whether the second argument is a variance, a standard deviation, or a precision.
All distributions, with the exception of binary and uniform, can have the optional arguments of
LOWER= and UPPER=, which specify a truncated density. See the section “ Truncation and
Censoring ” on page 4169 for more details.
Table 52.2
Valid Distributions
Distribution Name
a=
> ˛ , <
b=
> ˇ )
prob|p=
> p
)
(<
n=
> n , <
prob|p=
>
(<
location|loc|l=
>
df=
> ) p )
, <
scale|s=
> )
Definition beta distribution with shape parameters ˛ and ˇ binary (Bernoulli) distribution with probability of success p
. You can use the alias
bern
for this distribution.
binomial distribution with count n and probability of success p
Cauchy distribution with location
2 distribution with and scale degrees of freedom
MODEL Statement
F
4137
Table 52.2
(continued)
Distribution Name
ll )
df=
>
scale|s=
iscale|is=
)
)
shape|sp=
> a ,
scale|s=
shape|sp=
> a ,
iscale|is=
df=
>
df=
>
scale|s=
iscale|is=
)
)
)
)
, <
scale|s=
> s )
shape|sp=
> a ,
scale|s=
shape|sp=
> a ,
iscale|is=
)
prob|p=
> p )
)
)
)
shape|sp=
>
shape|sp=
> a ,
scale|s=
a ,
iscale|is=
)
)
Definition general log-likelihood function that you construct using SAS programming statements for single or multiple discrete variables. Also see the function
general
. The name
dlogden
function.
is an alias for this log transformation of a
2 distribution with degrees of freedom:
chisq
./ ,
expchisq
./
. You can use the alias log
. /
echisq
for this distribution.
log transformation of an exponential distribution with scale or inverse-scale parameter :
expon
./ , use the alias log
. /
eexpon expexpon
./
. You can for this distribution.
log transformation of a gamma distribution with shape a and scale or inverse-scale :
gamma
.a; /
,
You can use the alias log . /
egamma expgamma
.a; / .
for this distribution.
log transformation of an inverse with log . /
2 distribution degrees of freedom:
ichisq
./ ,
expichisq
./ . You can use the alias
eichisq
for this distribution.
log transformation of an inverse-gamma distribution with shape a and scale or inversescale :
expigamma
.a; /
igamma
.
.a; /
, log . /
You can use the alias
eigamma
for this distribution.
log transformation of a scaled inverse tion with
2 distribudegrees of freedom and scale parameter s
:
sichisq
./ , log
. /
You can use the alias
esichisq expsichisq
./
.
for this distribution.
exponential distribution with scale or inverse-scale parameter gamma distribution with shape inverse-scale a and scale or geometric distribution with probability p
4138
F
Chapter 52: The MCMC Procedure
Table 52.2
(continued)
Distribution Name
ll )
df=
>
)
shape|sp=
> a ,
scale|s=
shape|sp=
> a ,
iscale|is=
location|loc|l=
>
location|loc|l=
>
)
)
,
scale|s=
,
iscale|is=
)
)
location|loc|l=
> a , <
scale|s=
> b )
mean|m=
>
mean|m=
>
mean|m=
>
,
sd=
,
var|v=
,
prec=
)
)
)
Definition general log likelihood function that you construct using SAS programming statements for a single or multiple continuous variables. The argument ll is an expression for the log of the distribution. If there are multiple variables specified before the tilde in a MODEL, PRIOR, or HYPERPRIOR statement, ll is interpreted as the log of the joint distribution for these variables. Note that in the
MODEL statement, the response variable specified before the tilde is just a place holder and is of no consequence; the variable must have appeared in the construction of ments.
general
ll in the programming state-
( constant ) is equivalent to a uniform distribution on the real line. You can use the alias
logden
for this distribution.
inverse
2 distribution with degrees of freedom inverse-gamma distribution with shape a and scale or inverse-scale
Laplace distribution with location and scale or inverse-scale exponential
. This is also known as the double distribution. You can use the alias
dexpon
for this distribution.
logistic distribution with location a and scale b log-normal distribution with mean and standard deviation or variance or precision can use the aliases
lognormal
or
lnorm
distribution.
. You for this
n=
> n
, <
prob|p=
> p
)
mean|m=
>
mean|m=
>
mean|m=
>
,
,
sd= var|v=
)
,
prec=
shape|sp=
> a
, <
mean|m=
> )
)
)
scale|s=
> b
) negative binomial distribution with count n and probability of success p . You can use the alias
nb
for this distribution.
normal (Gaussian) distribution with mean standard deviation or variance or precision and
. You can use the aliases
gaussian
,
norm
, or
n
for this distribution.
Pareto distribution with shape a and scale b
Poisson distribution with mean
Table 52.2
(continued)
Distribution Name
df=
> , <
scale|s=
> s
)
mean|m=
>
mean|m=
>
mean|m=
>
,
sd=
,
var|v=
,
prec=
, <
df=
>
, <
df=
>
, <
df=
>
)
)
)
left|l=
> a , <
right|r=
> b )
mean|m=
> , <
iscale|is=
> )
; c;
)
PARMS Statement
F
4139
Definition scaled inverse
2 distribution with freedom and scale parameter s degrees of t distribution with mean variance or precision
, standard deviation or
, and degrees of freedom uniform distribution with range a and b . You can use the alias
unif
for this distribution.
Wald distribution with mean parameter verse scale parameter and in-
. This is also known as the Inverse Gaussian distribution. You can use the alias
igaussian
for this distribution.
Weibull distribution with location (threshold) parameter , shape parameter c , and scale parameter
.
PARMS
name | ( name-list ) < = > number < name | ( name-list ) <= > number . . . > < /
NORMAL
|
T
< (df) > |
UDS
>
;
The PARMS statement lists the names of the parameters in the model and specifies optional initial values for these parameters. Multiple PARMS statements are allowed. Each PARMS statement defines a block of parameters, and the blocked Metropolis algorithm updates the parameters in each
block simultaneously. See the section “ Blocking of Parameters ” on page 4148 for more details.
PROC MCMC generates missing initial values from the prior distributions whenever needed, as long as they are the standard distributions and not the functions
or
Every parameter in the PARMS statement must have a corresponding prior distribution in the PRIOR statement. The program exits if the one-to-one requirement is not satisfied.
The optional arguments give you control over different samplers explicitly for that block of parameters.
The normal proposal distribution in the random walk Metropolis is the default. You can also choose a t
-distribution with df degrees of freedom. If df
> 100 , the normal distribution is used instead.
The user defined sampler (UDS, see the section “ UDS Statement ” on page 4143) option allows
you to implement a new sampler for any of the parameters in the block. PROC MCMC does not use the Metropolis sampler on these parameters and incorporates your sampler to draw posterior samples. This can sometimes greatly improve the convergence and mixing of the Markov chain. This functionality is for advanced users, and you should proceed with caution.
4140
F
Chapter 52: The MCMC Procedure
PREDDIST
< ’label’ >
OUTPRED
=SAS-data-set <
NSIM
=n > <
COVARIATES
=SAS-data-set >
<
STATISTICS
=options >
;
The PREDDIST statement creates a new SAS data set that contains random samples from the posterior predictive distribution of the response variable. The posterior predictive distribution is the distribution of unobserved observations (prediction) conditional on the observed data. Let y be the observed data, X be the covariates, be the parameter, and y pred be the unobserved data. The posterior predictive distribution is defined to be the following: p.
y pred j y
;
X
/ D
D
Z p.
y pred
; j y
;
X
/d
Z p.
y pred j
; y ; X /p.
j y ; X /d
Given the assumption that the observed and unobserved data are conditional independent given the posterior predictive distribution can be further simplified as the following:
, p.
y pred j y ; X /
D
Z p.
y pred j
/p.
j y ; X /d
The posterior predictive distribution is an integral of the likelihood function to the posterior distribution p.
y pred j / with respect p.
j y / . The PREDDIST statement generates samples from a posterior predictive distribution based on draws from the posterior distribution of .
The PREDDIST statement works only on response variables that have standard distributions, and it does not support either the GENERAL or DGENERAL functions. Multiple PREDDIST statements can be specified, and an optional label (specified as a quoted string) helps identify the output.
The following list explains specifications in the PREDDIST statement:
COVARIATES=
SAS-data-set names the SAS data set that contains the sets of explanatory variable values for which the predictions are established. This data set must contain data with the same variable names as are used in the likelihood function. If you omit the COVARIATES= option, the DATA= data set specified in the PROC MCMC statement is used instead.
NSIM=
n specifies the number of simulated predicted values. By default, NSIM= uses the
option value specified in the PROC MCMC statement.
OUTPRED=
SAS-data-set creates an output data set to contain the samples from the posterior predictive distribution. The output variable names are listed as resp_1
– resp
_ m , where resp variable and m is the number of observations in the COVARIATES= data set in the PREDDIST statement. If the COVARIATES= data set is not specified, m is the name of the response is the number of observations in the DATA= data set specified in the PROC statement.
PRIOR/HYPERPRIOR Statement
F
4141
STATISTICS
< (global-stats-options) >
= NONE | ALL |
stats-request
STATS
< (global-stats-options) >
= NONE | ALL |
stats-request specifies options for calculating posterior statistics. This option works identically to the
option in the PROC statement. By default, this option takes the specification of the
option in the PROC MCMC statement.
For an example that uses the PREDDIST statement, see “ Posterior Predictive Distribution ” on
page 4185.
PRIOR
parameter-list distribution
;
HYPERPRIOR
HYPER
parameter-list parameter-list distribution
;
distribution
;
The PRIOR statement is used to specify the prior distribution of the model parameters. You must specify a single parameter or a list of parameters, a tilde ( ), and then a distribution with its parameters. Multiple
statements are allowed for defining models with multiple independent prior components. The log of the prior is the sum of the log prior values from each of the PRIOR
statements. See the section “ MODEL Statement ” on page 4136 for the names of the standard
distributions and the section “ Standard Distributions ” on page 4155 for density specification.
The PRIOR statements are processed twice at every Markov chain simulation—that is, twice per pass through the data set. The statements are called at the first and the last observation of the data set.
This is the same as how the
and
statements are processed.
The
statement is internally treated the same as the PRIOR statement. It provides a notational convenience in case you wish to fit a multilevel hierarchical model. It is used to specify the hyperprior distribution of the prior distribution parameters. The log of the hyperprior is the sum of the log hyperprior values from each of the
statements.
If you want to specify a multilevel hierarchical model, you can use either a
or a
statement as if it were a hyper-HYPERPRIOR statement. Your model can have as many hierarchical levels as desired.
This section lists the programming statements available in PROC MCMC to compute the priors and log-likelihood functions. This section also documents the differences between programming statements in PROC MCMC and programming statements in the DATA step. The syntax of programming statements used in PROC MCMC is identical to that used in the NLMIXED procedure
(see Chapter 61, “ The NLMIXED Procedure ”) and the MODEL procedure (see Chapter 18, “ The
SAS/ETS User’s Guide ),). Most of the programming statements that can
4142
F
Chapter 52: The MCMC Procedure be used in the DATA step can also be used in PROC MCMC. Refer to
SAS Language Reference:
Dictionary for a description of SAS programming statements.
There are also a number of unique functions in PROC MCMC that calculate the log density of
For the list of matrix-based functions that is supported in PROC MCMC, see the section “ Matrix
Functions in PROC MCMC ” on page 4176.
The following are valid statements:
ABORT;
CALL
name [ ( expression [, expression . . . ] ) ]
;
DELETE;
DO
[ variable = expression
[
TO
expression] [
BY
expression]
[
]
[, expression [
WHILE
TO
expression] [ expression ] [
BY
expression ] . . . ]
UNTIL
expression ]
;
END;
GOTO
statement_label
;
IF
expression
;
IF
expression
THEN
program_statement
;
ELSE
program_statement
;
variable = expression
;
variable + expression
;
LINK
statement_label
;
PUT
[ variable] [=] [...]
;
RETURN;
SELECT[(
expression
)];
STOP;
SUBSTR(
variable, index, length
)=
expression
;
WHEN
(expression) program_statement
;
OTHERWISE
program_statement
;
For the most part, the SAS programming statements work the same as they do in the DATA step, as documented in SAS Language Reference: Concepts . However, there are several differences:
The ABORT statement does not allow any arguments.
The DO statement does not allow a character index variable. Thus
do i = 1,2,3;
is supported; however, the following statement is not supported:
do i = 'A','B','C';
UDS Statement
F
4143
The PUT statement, used mostly for program debugging in PROC MCMC (see the section
“ Handling Error Messages ” on page 4193), supports only some of the features of the DATA
step PUT statement, and it has some features that are not available with the DATA step PUT statement:
– The PROC MCMC PUT statement does not support line pointers, factored lists, iteration factors, overprinting, _INFILE_, _OBS_, the colon (:) format modifier, or “$”.
–
The PROC MCMC PUT statement does support expressions, but the expression must be enclosed in parentheses. For example, the following statement displays the square root of x:
put (sqrt(x));
The WHEN and OTHERWISE statements enable you to specify more than one target statement.
That is, DO/END groups are not necessary for multiple statement WHENs. For example, the following syntax is valid:
select; when (exp1) stmt1; stmt2; when (exp2) stmt3; stmt4; end;
You should avoid defining variables that begin with an underscore (_). They might conflict with internal variables created by PROC MCMC. The
statement must come after any SAS programming statements that define or modify terms used in the construction of the log likelihood.
UDS
subroutine-name (subroutine-argument-list)
;
UDS stands for user defined sampler. The UDS statement allows you to use a separate algorithm, other than the default random walk Metropolis, to update parameters in the model. The purpose of the UDS statement is to give you a greater amount of flexibility and better control over the updating schemes of the Markov chain. Multiple UDS statements are allowed.
For the UDS statement to work properly, you have to do the following: write a subroutine by using PROC FCMP (see the FCMP Procedure in the Base SAS Procedures
Guide ) and save it to a SAS catalog (see the example in this section). The subroutine must update some parameters in the model. These are the UDS parameters. The subroutine is called the UDS subroutine.
declare any UDS parameters in the
statement with a sampling option, as in < / UDS >
(see the section “ PARMS Statement ” on page 4139).
4144
F
Chapter 52: The MCMC Procedure specify the prior distributions for all UDS parameters, using the
statements.
N
OTE
: All UDS parameters must appear in three places: the UDS statement, the
statement, and the
statement. Otherwise, PROC MCMC exits.
To obtain a valid Markov chain, a UDS subroutine must update a parameter from its full posterior conditional distribution and not the posterior marginal distribution. The posterior conditional is something that you need to provide. This conditional is implicitly based on a prior distribution.
PROC MCMC has no means to verify that the implied prior in the UDS subroutine is the same as the prior that you specified in the
statement. You need to make sure that the two distributions agree; otherwise, you will get misleading results.
The priors in the
statements do not directly affect the sampling of the UDS parameters. They could affect the sampling of the other parameters in the model, which, in turn, changes the behavior of the Markov chain. You can see this by noting cases where the hyperparameters of the UDS parameters are model parameters; the priors should be part of the posterior conditional distributions of these hyperparameters, and they cannot be omitted.
Some additional information is listed to help you better understand the UDS statement:
Most features of the SAS programming language can be used in subroutines processed by
PROC FCMP (see the FCMP Procedure in the Base SAS Procedures Guide ).
The UDS statement does not support FCMP functions—a FCMP function returns a value, while a subroutine does not. A subroutine updates some of its subroutine arguments. These arguments are called OUTARGS arguments.
The UDS parameters cannot be in the same block as other parameters. The optional argument
< / UDS > in the
statement prevents parameters that use the default Metropolis from being mixed with those that are updated by the UDS subroutines.
You can put all the UDS parameters in the same
statement or have a separate UDS statement for each of them.
The same subroutine can be used in multiple UDS statements. This feature comes in handy if you have a generic sampler that can be applied to different parameters.
PROC MCMC updates the UDS parameters by calling the UDS subroutines directly. At every iteration, PROC MCMC first samples parameters that use the Metropolis algorithm, then the
UDS parameters. Sampling of the UDS parameters proceeds in the order in which the UDS statements are listed.
A UDS subroutine accepts any symbols in the program as well as any input data set variables as its arguments.
Only the OUTARGS arguments in a UDS subroutine are updated in PROC MCMC. You can modify other arguments in the subroutine, but the changes are not global in the procedure.
If a UDS subroutine has an argument that is a SAS data set variable, PROC MCMC steps through the data set while updating the UDS parameters. The subroutine is called once per observation in the data set for every iteration.
UDS Statement
F
4145
If a UDS subroutine does not have any arguments that are data set variables, PROC MCMC does not access the data set while executing the subroutine. The subroutine is called once per iteration.
To reduce the overhead in calling the UDS subroutine and accessing the data set repeatedly, you might consider reading all the input data set variables into arrays and using the arrays as the
subroutine arguments. See the section “ BEGINCNST/ENDCNST Statement ” on page 4133
about how to use the
and
statements to store data set variables.
An Example that Uses the UDS Statement
Suppose that you are interested in modeling normal data with conjugate prior distributions. The data are as follows:
title 'An Example that uses the UDS Statement'; data a; input y @@; i = _n_; datalines;
-0.651
-5.754
;
17.435
-5.002
-5.943
-2.545
-2.543 -10.444
-1.743
0.998
The likelihood for each observation is as follows: f .y
i j ; / D .; var
D
2
/
The prior distributions on and
2 are as follows:
.
j
.
2 j
0
;
2
0
/
0
;
2
0
/
D
D
.
0
; var D f si
2
.
shape
D
2
0
/
0
; scale
D
2
0
/ where f si
2 is the density function for a scaled inverse chi-square distribution. To sample without using any UDS statements, you can use the following program: and
2
proc mcmc data=a seed=17; parm mu; parm s2; begincnst; mu0 = 0; t0 = 20; nu0 = 10; s0 = 10; endcnst; prior mu ~ normal(mu0, var=t0); prior s2 ~ sichisq(nu0, s0); model y ~ normal(mu, var = s2); run;
4146
F
Chapter 52: The MCMC Procedure
This is a case where the full posterior conditional distribution of
It is also a normal distribution: p.
j
2
; y/ D
0
@
0
2
0
1
2
0
C
C n N
2 n
2
;
1
1
1
0
C n
2
A given
2 and y has a closed form.
You can define a subroutine, muupdater , which generates a random normal sample from the posterior conditional distribution described previously.
proc fcmp outlib=sasuser.funcs.uds; subroutine muupdater(mu, s2, mu0, t0, n, sumy); outargs mu; sigma2 = 1 / (1/t0 + n/s2); mean = (mu0/t0 + sumy/s2) * sigma2; mu = rand("normal", mean, sqrt(sigma2)); endsub; run;
The subroutine is saved in the OUTLIB= library. The declaration of any subroutine begins with a SUBROUTINE statement and ends with an ENDSUB statement. The OUTARGS statement in the subroutine indicates that mu is updated. Others, such as sigma2
, mu0
, and so on, are arguments that are needed in the full conditional distribution. Here the rand and sqrt are two of the many SAS functions that you can use.
You specify a CMPLIB option to let SAS search each of the catalogs that are specified in the option for a package that contains muupdater .
options cmplib=sasuser.funcs;
To use the subroutine in the UDS statement, you can use the following statements:
proc mcmc data=a seed=17;
UDS muupdater(mu, s2, mu0, t0, n, sumy); parm mu /uds; parm s2; begincnst; mu0 = 0; t0 = 20; nu0 = 10; s0 = 10; n = 10; if i eq 1 then sumy = 0; sumy = sumy + y; call streaminit(1); endcnst; prior mu ~ normal(mu0, var=t0); prior s2 ~ sichisq(nu0, s0); model y ~ normal(mu, var = s2); run;
These statements are very similar to the previous program. The differences are the UDS statement, the < / UDS > option in the
statement, and a few lines that computes the values of sumy and n
.
Details: MCMC Procedure
F
4147
The symbol sumy is the sum of y . The value is obtained by taking advantage of the
and
ENDCNST statements. See the example in the section “ BEGINCNST/ENDCNST Statement ”
on page 4133. The symbol n is the sample size in the data set.
The CALL STREAMINIT routine ensures that the RAND function in ducible stream of random numbers. The
muupdater creates a reprooption specifies a seed for the random number generator in PROC MCMC, which does not control the random number generator in the RAND function in the subroutine. You need to set both to reproduce the same stream of Markov chain samples.
The two programs produce different but similar numbers (results not shown) for the posterior distributions of and
2
.
For a more realistic example that uses the UDS statement, see “ Example 52.11: Implement a New
Sampling Algorithm ” on page 4296.
PROC MCMC uses a random walk Metropolis algorithm to obtain posterior samples. For details
on the Metropolis algorithm, see the section “ Metropolis and Metropolis-Hastings Algorithms ” on
page 150. For the actual implementation details of the Metropolis algorithm in PROC MCMC, such
Proposal Distribution ” on page 4150. By default, PROC MCMC assumes that all observations in the
data set are independent, and the logarithm of the posterior density is calculated as follows: log
.p.
j y // D log
.. // C n
X log
.f .y
i j // i D 1 where is a parameter or a vector of parameters. The term log .. // is the sum of the log of the prior densities specified in the
the log likelihood specified in the and
statement. The statements. The term log
.f .y
i j //
is statement specifies the log likelihood for a single observation in the data set.
The statements in PROC MCMC are in many ways like DATA step statements; PROC MCMC evaluates every statement in order for each observation. The procedure cumulatively adds the log likelihood for each observation. Statements between the
and
statements are evaluated only at the first and the last observations. At the last observation, the log of the prior and hyperprior distributions is added to the sum of the log likelihood to obtain the log of the posterior distribution.
With multiple
statements (multiple blocks of parameters), PROC MCMC updates each block of parameters while holding the others constants. The procedure still steps through all of the programming statements to calculate the log of the posterior distribution, given the current or
4148
F
Chapter 52: The MCMC Procedure the proposed values of the updating block of parameters. In other words, the procedure does not calculate the conditional distribution explicitly for each block of parameters, and it uses the full joint distribution in the Metropolis step for every block update. If you wish to model dependent data—that is, log .f .
y j
//
¤
P i log .f .y
i j
// —you can use the PROC option
“ Modeling Joint Likelihood ” on page 4181 for more details.
In a multivariate parameter model, if all k parameters are proposed with one joint distribution q.
j
/ , acceptance or rejection would occur for all of them. This can be rather inefficient, especially when parameters have vastly different scales. A way to avoid this difficulty is to allocate the into d k parameters blocks and update them separately. The
statement is used to specify model parameters.
It also puts parameters in separate blocks, and each block of parameters is updated sequentially in the procedure.
Suppose that you wish to sample from a multivariate distribution with probability density function p.
j y / where
D f
1
;
2
; : : : ; k g :
Now suppose that these k parameters are separated into d blocks—for example, nonempty subset of the p.
j x
/ f i g
D f d
.z/ where
, and where each i z D f z
1
; z
2
; : : : ; z d g
, where each is contained in one and only one z j z j contains a
. In the MCMC context, the z
’s are blocks of parameters. In the blocked algorithm, a proposal is composed of several parts. Instead of proposing a simultaneous move for all the ’s, a proposal is made for the i
’s in z
1 only, then for the i
’s in z
2
, and so on for d subproposals. Any accepted proposal can involve any number of the blocks moving. Not necessarily all of the parameters move at once as in the all-at-once
Metropolis algorithm.
Formally, the blocked Metropolis algorithm is as follows. Let w j block z j w j
.
and let q j
.
j w j
/ be the collection of i that are in be a symmetric multivariate distribution centered at the current values of
1. Let t
D
0 . Choose points for all w j t
. This can be an arbitrary point as long as p.w
t j j y / > 0 .
2. For j
D
1; ; d : a) Generate a new sample, w j;new
, using the proposal distribution q j
.
j w j t
/ .
b) Calculate the following quantity: r D min
( p.w
j;new j w t
1 p.w
t j j w t
1
;
; ; w t j 1
; w
; w t j 1 t 1 j C 1
; w t C 1 j 1
;
;
; w t d
; y /
; 1
; w t d
; y /
)
: c) Sample u from the uniform distribution
U.0; 1/
.
d) Set w t C 1 j
D w j;new if r < a
; w t C 1 j
D w t j otherwise.
3. Set t D t C 1
. If t < T
, the number of desired samples, go back to Step 2; otherwise, stop.
With PROC MCMC, you can sample all parameters simultaneously by putting them all in a single
PARMS statement, you can sample parameters individually by putting each parameter in its own
Samplers
F
4149
PARMS statement, or you can sample certain subsets of parameters together by grouping each subset in its own PARMS statements. For example, if the model you are interested in has five parameters, alpha
, beta
, gamma
, phi
, sigma
, the all-at-once strategy is as follows:
parms alpha beta gamma phi sigma;
The one-at-a-time strategy is as follows:
parms alpha; parms beta; parms gamma; parms phi; parms sigma;
A two-block strategy could be as follows:
parms alpha beta gamma; parms phi sigma;
One of the greatest challenges in MCMC sampling is achieving good mixing of the chains—the chains should quickly traverse the support of the stationary distribution. A number of factors determine the behavior of a Metropolis sampler; blocking is one of them, so you want to be extra careful when it comes to choosing a good design. Generally speaking, forming blocks of parameters has its advantages, but it is not true that the larger the block the faster the convergence.
When simultaneously sampling a large number of parameters, the algorithm might find it difficult to achieve good mixing. As the number of parameters gets large, it is much more likely to have
(proposal) samples that fall well into the tails of the target distribution, producing too small a test ratio. As a result, few proposed values are accepted and convergence is slow. On the other hand, when sampling each parameter individually, the chain might mix far too slowly because the conditional distributions (of i given all other ’s) might be very “narrow.” Hence, it takes a long time for the chain to explore fully that dimension alone. There are no theoretical results that can help determine an optimal “blocking” for an arbitrary parametric model. A rule followed in practice is to form small groups of correlated parameters that belong to the same context in the formulation of the model. The best mixing is usually obtained with a blocking strategy somewhere between the all-at-once and one-at-a-time strategies.
This section describes the sampling methods used in PROC MCMC. Each block of parameters is classified by the nature of the prior distributions. “Continuous” means all priors of the parameters in the same block are continuous distribution. “Discrete” means all priors are discrete. “Mixed” means that some parameters are continuous and others are discrete. Parameters that have binary priors are treated differently, as indicated in the table. MVN stands for the multivariate normal distribution, and MVT is short for the multivariate t -distribution.
Blocks continuous
Default Method
MVN
Alternative Method
MVT
4150
F
Chapter 52: The MCMC Procedure
Blocks discrete (other than binary) mixed binary (single dimensional) binary (multi-dimensional)
Default Method binned MVN
MVN inverse CDF independence sampler
Alternative Method binned MVT or symmetric geometric
MVT
For a block of continuous parameters, PROC MCMC uses a multivariate normal distribution as the default proposal distribution. In the tuning phase, the procedure finds an optimal scale c covariance matrix
†
.
and a tuning
For a discrete block of parameters, PROC MCMC uses a discretized multivariate normal distribution as the default proposal distribution. The scale c and covariance matrix † are tuned. Alternatively, you can use an independent symmetric geometric proposal distribution. The density has form p.1 p/ j
2.1 p/ j and has variance
.2 p/.1 p/ p
2
. In the tuning phase, the procedure finds an optimal proposal probability p for every parameter in the block.
You can change the proposal distribution, from the normal to a t -distribution. You can either use the PROC option
The
or
statement option < / T( df ) > to make the change.
t -distributions have thicker tails, and they can propose to the tail areas more efficiently than the normal distribution. It can help with the mixing of the Markov chain if some of the parameters have
a skewed tails. See “ Example 52.4: Nonlinear Poisson Regression Models ” on page 4229. The
independence sampler (see the section “ Independence Sampler ” on page 153) is used for a block
of binary parameters. The inverse CDF method is used for a block that consists of a single binary parameter.
One key factor in achieving high efficiency of a Metropolis-based Markov chain is finding a good proposal distribution for each block of parameters. This process is referred to as tuning. The tuning phase consists of a number of loops. The minimum number of loops is controlled by the option
MINTUNE= , with a default value of 2. The option MAXTUNE=
controls the maximum number of tuning loops, with a default value of 24. Each loop lasts for
iterations, where by default
500. At the end of every loop, PROC MCMC examines the acceptance probability for each block. The acceptance probability is the percentage of
proposals that have been accepted.
If the probability falls within the acceptance tolerance range (see the section “ Scale Tuning ” on
page 4151), the current configuration of c
/ † or p is kept. Otherwise, these parameters are modified before the next tuning loop.
Continuous Distribution: Normal or t -Distribution
A good proposal distribution should resemble the actual posterior distribution of the parameters.
Large sample theory states that the posterior distribution of the parameters approaches a multivariate normal distribution (see
Gelman et al. 2004 , Appendix B, and
Schervish 1995 , Section 7.4). That is
Tuning the Proposal Distribution
F
4151 why a normal proposal distribution often works well in practice. The default proposal distribution in
PROC MCMC is the normal distribution: q j
.
new j t
/ D
MVN
.
new j t
; c
2
†
/
. As an alternative, you can choose a multivariate t -distribution as the proposal distribution. It is a good distribution to use if you think that the posterior distribution has thick tails and a t -distribution can improve
the mixing of the Markov chain. See “ Example 52.4: Nonlinear Poisson Regression Models ” on
page 4229.
Scale Tuning
The acceptance rate is closely related to the sampling efficiency of a Metropolis chain. For a random walk Metropolis, high acceptance rate means that most new samples occur right around the current data point. Their frequent acceptance means that the Markov chain is moving rather slowly and not exploring the parameter space fully. On the other hand, a low acceptance rate means that the proposed samples are often rejected; hence the chain is not moving much. An efficient Metropolis sampler has an acceptance rate that is neither too high nor too low. The scale c in the proposal distribution q.
j / effectively controls this acceptance probability.
Roberts, Gelman, and Gilks ( 1997
showed that if both the target and proposal densities are normal, the optimal acceptance probability
for the Markov chain should be around 0.45 in a single dimensional problem, and asymptotically approaches 0.234 in higher dimensions. The corresponding optimal scale is
2:38
, which is the initial scale set for each block.
Due to the nature of stochastic simulations, it is impossible to fine-tune a set of variables such that the Metropolis chain has the exact desired acceptance rate. In addition,
Roberts and Rosenthal ( 2001 )
empirically demonstrated that an acceptance rate between 0.15 and 0.5 is at least 80% efficient, so there is really no need to fine-tune the algorithms to reach acceptance probability that is within small tolerance of the optimal values. PROC MCMC works with a probability range, determined by the
PROC options
˙
ACCEPTTOL . The default value of
is a function of the number of parameters in the model, as outlined in default value of
is
Roberts, Gelman, and Gilks ( 1997 ). The
0:075
. If the observed acceptance rate in a given tuning loop is less than the lower bound of the range, the scale is reduced; if the observed acceptance rate is greater than the upper bound of the range, the scale is increased. During the tuning phase, a scale parameter in the normal distribution is adjusted as a function of the observed acceptance rate and the target acceptance rate. The following updating scheme is used in PROC MCMC
1
: c new
D c cur
ˆ
ˆ
1
.p
opt
=2/
1
.p
cur
=2/ where c cur is the current scale, probability.
p cur is the current acceptance rate, p opt is the optimal acceptance
Covariance Tuning
To tune a covariance matrix, PROC MCMC takes a weighted average of the old proposal covariance matrix and the recent observed covariance matrix, based on
samples in the current loop. The
1
acceptance probability and scale in a random walk Metropolis is p D 2ˆ I c=2
, where c is the scale, p is the acceptance rate,
ˆ is the CDF of a standard normal, and
I E f samples. This relationship determines the updating scheme, with I
Œ.f
0
.x/=f .x//
2
, f .x/ is the density function of being replaced by the identity matrix to simplify calculation.
4152
F
Chapter 52: The MCMC Procedure
option determines how much weight is put on the recently observed covariance matrix.
The formula used to update the covariance matrix is as follows:
COVnew
D w COVcur
C .1
w
/
COVold
There are two ways to initialize the covariance matrix:
The default is an identity matrix multiplied by the initial scale of
2:38
(controlled by the PROC option
SCALE= ) and divided by the square root of the number of estimated parameters in
the model. It can take a number of tuning phases before the proposal distribution is tuned to its optimal stage, since the Markov chain needs to spend time learning about the posterior covariance structure. If the posterior variances of your parameters vary by more than a few orders of magnitude, if the variances of your parameters are much different from 1, or if the posterior correlations are high, then the proposal tuning algorithm might have difficulty with forming an acceptable proposal distribution.
Alternatively, you can use a numerical optimization routine, such as the quasi-Newton method, to find a starting covariance matrix. The optimization is performed on the joint posterior distribution, and the covariance matrix is a quadratic approximation at the posterior mode. In some cases this is a better and more efficient way of initializing the covariance matrix. However, there are cases, such as when the number of parameters is large, where the optimization could fail to find a matrix that is positive definite. In that case, the tuning covariance matrix is reset to the identity matrix.
A side product of the optimization routine is that it also finds the estimates with respect to the posterior distribution.
maximum a posteriori (MAP)
The MAP estimates are used as the initial values of the Markov chain.
If any of the parameters are discrete, then the optimization is performed conditional on these discrete parameters at their respective fixed initial values. On the other hand, if all parameters are continuous, you can in some cases skip the tuning phase (by setting
MAXTUNE=0 ) or the burn-in phase (by
setting
Discrete Distribution: Symmetric Geometric
By default, PROC MCMC uses the normal density as the proposal distribution in all Metropolis random walks. For parameters that have discrete prior distributions, PROC MCMC discretizes proposed samples. You can choose an alternative symmetric geometric proposal distribution by specifying the option
The density of the symmetric geometric proposal distribution is as follows: p g
.1
2.1
p g
/ j p g
/ j where the symmetry centers at
2
D
.2
p g
/.1
p
2 g p g
/
. The distribution has a variance of
Initial Values of the Markov Chains
F
4153
Tuning for the proposal p g uses the following formula: new cur
D
ˆ
ˆ
1
.p
opt
=2/
1
.p
cur
=2/ where new is the standard deviation of the new proposal geometric distribution, deviation of the current proposal distribution, p cur is the standard opt is the target acceptance probability, and p cur is the current acceptance probability for the discrete parameter block.
The updated p g is the solution to the following equation that is between 0 and 1 : s
.2
p g
/.1
p
2 g p g
/
D cur
ˆ
ˆ
1
.p
opt
=2/
1
.p
cur
=2/
Binary Distribution: Independence Sampler
Blocks consisting of a single parameter with a binary prior do not require any tuning; the inverse-CDF method applies. Blocks that consist of multiple parameters with binary prior are sampled by using
an independence sampler with binary proposal distributions. See the section “ Independence Sampler ”
on page 153. During the tuning phase, the success probability p of the proposal distribution is taken to be the probability of acceptance in the current loop. Ideally, an independence sampler works best if the acceptance rate is
100
%, but that is rarely achieved. The algorithm stops when the probability of success exceeds the
TARGACCEPTI= value , which has a default value of
0:6
.
You can assign initial values to any parameters. To assign initial values, you can either use the
statements or use programming statements within the
and
statements. For
the latter approach, see the section “ BEGINCNST/ENDCNST Statement ” on page 4133.
When parameters have missing initial values, PROC MCMC tries to generate them from the respective
page 4155. PROC MCMC either uses the mode from the prior distribution or draws a random number from it. For distributions that do not have modes, such as the uniform distribution, PROC MCMC uses the mean instead. In general, PROC MCMC avoids using starting values that are close to the boundary of support of the prior distribution. For example, the exponential prior has a mode at
0
, and
PROC MCMC starts an initial value at the mean. This avoids some potential numerical problems. If you use the
or
functions in the
statements, you must provide initial values for those parameters.
If you use the optimization option
PROPCOV , PROC MCMC starts the tuning at the optimized
values. The procedure overwrites the initial values that you provided unless you use the option
4154
F
Chapter 52: The MCMC Procedure
In general, you cannot alter the values of any model parameters in PROC MCMC. For example, the following assignment statement produces an error:
parms alpha; alpha = 27;
This restriction prevents incorrect calculation of the posterior density—assignments of parameters in the program would override the parameter values generated by the procedure and lead to a constant value of the density function.
However, you can modify parameter values and assign initial values to parameters within the block defined by the
and
statements. The following syntax is allowed:
parms alpha; begincnst; alpha = 27; endcnst;
The initial value of alpha is
27
. Assignments within the BEGINCNST/ENDCNST block override initial values specified in the
statement. For example, with the following statements, the
Markov chain starts at alpha
D
27 , not 23 .
parms alpha 23; begincnst; alpha = 27; endcnst;
This feature enables you to systematically assign initial values. Suppose that z is an array parameter of the same length as the number of observations in the input data set. You want to start the Markov chain with each z i statements set z i having a different value depending on the data set variable
D j y j for the first half of the observations and z i
D 2:3 y
. The following for the rest:
/* a rather artificial input data set. */ data inputdata; do ind = 1 to 10; y = rand('normal'); output; end; run; proc mcmc data=inputdata; array z[10]; begincnst; if ind <= 5 then z[ind] = abs(y); else z[ind] = 2.3; endcnst; parms z:; prior z: ~ normal(0, sd=1); model general(0); run;
Standard Distributions
F
4155
Elements of z are modified as PROC MCMC executes the programming statements between the
and
statements. This feature could be useful when you use the
function and you find that the
statements are too cumbersome for assigning starting values.
through
show all densities that PROC MCMC recognizes. These densities can be used in the
statements. See the section “ Using
Density Functions in the Programming Statements ” on page 4167 for information about how to
use distributions in the programming statements. To specify an arbitrary distribution, you can use the functions
and
DGENERAL . See the section “ Specifying a New Distribution ” on
page 4166 for more details. See the section “ Truncation and Censoring ” on page 4169 for tips on
how to work with truncated distributions and censoring data.
Table 52.4
Beta Distribution
PROC specification density parameter restriction range mean variance mode random number beta
( a
, b
)
.a
C b/
.a/.b/ a 1
.1
/ b 1 a > 0
,
8
ˆ
ˆ
ˆ
ˆ
ˆ
< b > 0
Œ0; 1
Œ0; 1/ when when a a
D 1; b D 1
D 1; b ¤ 1
ˆ
ˆ
ˆ
ˆ
ˆ
:
.0; 1
.0; 1/ when a ¤ otherwise a a C b ab
.a
C b/
2
.a
C b C 1/
8 a 1
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ a C b 2
0 and
1
ˆ
ˆ
ˆ
ˆ
ˆ
<
0
1; b D 1 a > 1; b > 1 a < 1; b < 1
( a < 1; b 1 a D 1; b > 1
( a 1; b < 1
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
:
1 does not exist uniquely a a > 1; b
D b D 1
D 1 if min .a; b/ > 1
.a; b/ < 1
, see ( Atkinson and Whittaker 1976 ) and ( Atkinson 1979 ); if min
.a; b/ < 1 and max
.a; b/ > 1 method; if a
D b
D
a D 1 or
1 , uniform random variable b D 1
, inversion
4156
F
Chapter 52: The MCMC Procedure
Table 52.5
Binary Distribution
PROC specification density parameter restriction range mean variance mode random number binary ( p ) p .1
p/
1
0
8 p
ˆ
ˆ
< f 0 g f 1 g
ˆ
ˆ
: f
0; 1 g
1 when p D 0 when p D 1 otherwise round
.p/ p.1
( f 1 g p/ when p D 1 f 0 g otherwise generate u uniform .0; 1/ . If u p ,
D
1 ; else,
D
0
Table 52.6
Binomial Distribution
PROC specification density binomial
( n
, p
)
!
n p .1
p/ n parameter restriction range mean variance mode n D 0; 1; 2; 0
2 f
0; ; n g b np c np.1
p/ b .n
C 1/p c p 1
Table 52.7
Cauchy Distribution
PROC specification density parameter restriction range mean variance mode random number cauchy ( a
, b
)
1 b
2 b
C . a/
2 b > 0
2 .
1 ; 1 / does not exist does not exist a generate u
1
; u
2 procedure until u
2
1 standard Cauchy, and uniform
.0; 1/
, let
C v
2
< 1
.
y D v D v=u
1
2u
2
1
. Repeat the is a draw from the
D a
C by
Standard Distributions
F
4157
Table 52.8
2
Distribution
PROC specification density parameter restriction range mean variance mode random number chisq ( )
1
.=2/2
=2
> 0
.=2/ 1
2 Œ0; 1 / if
D 2 e
;
=2
.0; 1 / otherwise
2
2 if 2 ; does not exist otherwise
2 is a special case of the gamma distribution: gamma
.=2; scale=
2/ is a draw from the
2 distribution
Table 52.9
Exponential
2
Distribution
PROC specification density parameter restriction range mode random number relationship to the tribution
2 disexpchisq ( )
1
.=2/2
=2 exp
. /
=2 exp
.
> 0 exp
. /=2/
2 .
1 ; 1 / log
./ generate nential
2 x
1
2
./
, and distribution
2
./ , log
. / exp
D log
.x
1
/ is a draw from the expo-
2
./
Table 52.10
Exponential Exponential Distribution
PROC specification density parameter restriction range mode random number relationship to the Expon distribution expexpon ( scale = b )
1 b exp
. / exp
.
b > 0 exp
. /=b/ expexpon ( iscale =
ˇ )
ˇ exp
. / exp
.
ˇ > 0 exp
. / ˇ/
2 .
log
.b/
1 ; 1 / same log
.1=ˇ/ generate x
1 expon .
scale= b/ , and
D log .x
1
/ is a draw from the exponential exponential distribution. Note that an exponential exponential distribution is not the same as the double exponential distribution.
expon
.b/ , log
. / expExpon
.b/
4158
F
Chapter 52: The MCMC Procedure
Table 52.11
Exponential Gamma Distribution
PROC specification density parameter restriction range mode random number relationship to the distribution
expgamma ( a , scale = b )
1 b a
.a/ e a exp
. e a > 0; b > 0
=b/
2 .
log .ab/
1 ; 1 / expgamma ( a ,
ˇ a
.a/ e a a > 0; ˇ > 0 same log .a=ˇ/ iscale = exp
. e ˇ/
ˇ ) generate x
1 gamma
.a; scale
D b/
, and from the exponential gamma distribution gamma
.a; b/ , log
. /
D expGamma
.a; b/ log
.x
1
/ is a draw
Table 52.12
Exponential Inverse
2
Distribution
PROC specification density parameter restriction range mode random number relationship to the distribution i
2 expichisq ( )
.
2
1
/2
=2
> 0 exp . =2/ exp . 1=.2
exp . ///
2 .
1 ; 1 / log ./ generate x
1 nential inverse i
2
./
, and
2 distribution
D log
.x
1
/ is a draw from the expoi
2
./ , log . / exp i
2
./
Table 52.13
Exponential Inverse-Gamma Distribution
PROC specification density parameter restriction range mode random number relationship to the distribution i expigamma ( a
, scale = b
) b a
.a/ exp . ˛ / exp . b= exp . // a > 0; b > 0 expigamma ( a
, iscale =
ˇ
)
ˇ
1
˛
.a/ exp . ˛ / exp .
a > 0; ˇ > 0
ˇ
1 exp
. /
/
2 .
log
1 ; 1 /
.a=b/ same log
.aˇ/ generate x
1 igamma
.a; scale
D b/
, and
D from the exponential inverse-gamma distribution log
.x
1
/ is a draw igamma
.a; b/ , log
. / eigamma
.a; b/
Standard Distributions
F
4159
Table 52.14
Exponential Scaled Inverse
2
Distribution
PROC specification density parameter restriction range mode random number relationship to the distribution si
2 expsichisq (
.
2
/
.
=2
2
/ s exp
, s )
. =2/ exp
. s
2
=.2
exp
. ///
> 0; s > 0
2 .
log
.s
2
/
1 ; 1 / generate x
1 si
2
./ , and exponential scaled inverse
2
D log .x
distribution
1
/ is a draw from the si
2
./ , log
. / exp si
2
./
Table 52.15
Exponential Distribution
PROC specification density parameter restriction range mean variance mode random number expon
( scale = b
)
1 b e
=b b > 0 b b
2
0
2 Œ0; 1 / expon
( iscale =
ˇ
ˇe
ˇ
ˇ > 0 same
1=ˇ
1=ˇ
2
0
) the exponential distribution is a special case of the gamma distribution: gamma
.1; scale
D b/ is a draw from the exponential distribution
Table 52.16
Gamma Distribution
PROC specification density parameter restriction range mean variance mode random number gamma ( a , scale =
1 b a
.a/ a 1
2 Œ0; 1 / erwise e a > 0; b > 0
=b if a b )
D 1 I .0; 1 / othab ab
2
.a
1/b if a 1
see ( McGrath and Irving 1973 )
gamma ( a , iscale =
ˇ a
.a/ a 1 e
ˇ
ˇ a > 0; ˇ > 0 same a=ˇ a=ˇ
2
.a
1/=ˇ if a 1
)
4160
F
Chapter 52: The MCMC Procedure
Table 52.17
Geometric Distribution
PROC specification density
2 parameter restriction range mean variance mode random number geo ( p ) p.1
p/
0 < p 1
(
2 f 0; 1; 2; : : : f 0 g round(
1 p p
)
1 p p
2 g
0
0 < p < 1 p D 1 based on samples obtained from a Bernoulli distribution with probability p until the first success
Table 52.18
Inverse
2
Distribution
PROC specification density parameter restriction range mean variance mode random number ichisq
( )
1
.=2/2
=2
> 0
.=2 C 1/ e
1=.2 /
2
.0;
1
/
1
2 if
> 2
2
. 2/
2
. 4/ if > 4
1
C 2 inverse igamma
2 is a special case of the inverse-gamma distribution:
.=2; scale
D 2/ is a draw from the inverse
2 distribution
2
The random variable is the total number of failures in an experiment before is not to be confused with another popular formulation, p.1
p/
1 the first success. This density function
, which counts the total number of trials until the first success.
Standard Distributions
F
4161
Table 52.19
Inverse-Gamma Distribution
PROC specification density parameter restriction range mean variance mode random number relationship to the gamma distribution igamma ( a , scale = b b a
.a/
.a
C 1/ e b=
) a > 0; b > 0 igamma ( a , iscale =
ˇ )
ˇ a
1
.a/
.a
C 1/ a > 0; ˇ > 0 e
1=ˇ
2 .0; 1 / same b a 1 if a > 1 b
2
.a 1/
2
.a 2/ b a C 1
1
ˇ .a 1/ generate x
1 the igamma .a; gamma iscale
.a;
D scale b/
D b/
, and distribution if
ˇ
2
1
.a 1/
2
.a 2/
1
ˇ .a
C 1/
D a > 1
1=x
1 is a draw from gamma
.a; scale
D b/ , 1= igamma
.a; iscale
D b/
Table 52.20
Laplace (Double Exponential) Distribution
PROC specification density parameter restriction range mean variance mode random number laplace ( a , scale = b )
1
2b e b > 0 j a j =b laplace ( a , iscale =
ˇ )
ˇ
2 e
ˇ j a j
ˇ > 0
2 .
a
2b
2
1 ; 1 / same a
2=ˇ
2 a a inverse CDF.
u
1
; u
2
D a
F . /
D
8
<
:
1
2
1 exp
1
2 exp a b
a b
< a a
: Generate b uniform log .u
2
/ .
.0; 1/
. If u
1
< 0:5; D a C b log
.u
2
/ I is a draw from the Laplace distribution else
4162
F
Chapter 52: The MCMC Procedure
Table 52.21
Logistic Distribution
PROC specification density parameter restriction range mean variance mode random number logistic ( a , b ) b exp
.
.
1 C exp
.
b a
/ b a
//
2 b > 0
2 .
1 ; 1 / a
2 b
2
3 a inverse CDF method with
F . / u uniform .0; 1/ , and D a
D b
1 C exp
.
log .1=u
a b
/
1
. Generate
1/ is a draw from the logistic distribution
Table 52.22
LogNormal Distribution
PROC specification density parameter restriction range mean lognormal s p
2 s > 0 exp
( , sd = s
)
.
log
/
2
2s
2 lognormal p
1
2v v > 0 exp
( , var = v
)
.
log
/
2
2v lognormal
1 q
2
> 0 exp
( , prec =
)
.
log
/
2
2 variance mode random number
2
.0;
1
/ exp
.
C s
2
=2/ exp .2.
C s
2
// exp exp
.2
C s
2
/
.
s
2
/ same exp exp exp
.
.2.
exp
.
generate x
1 normal lognormal distribution
.0; 1/ , and
C v=2/
.2
v/
C v//
D
C v/ exp .
C sx
1
/ same exp exp
.
exp .2.
exp
.
C 1=.2 //
.2
C
C
1= /
1= //
1= / is a draw from the
Standard Distributions
F
4163
Table 52.23
Negative Binomial Distribution
PROC specification density negbin ( n , p )
C n 1
!
p n
.1
p/ parameter restriction range mean variance mode random number n D
2
1; 2;
(
0 < p f 0; 1; 2; : : : g f
0 g round n.1 p/ p n.1 p/ p
2
8
<
0
: round
.n 1/.1 p/ p
1
0 < p < 1 p
D
1 n D 1 n > 1 generate x
1
gamma
.n; 1/
, and Poisson
.x
1
.1
p/=p/
Table 52.24
Normal Distribution
PROC specification density parameter restriction range mean variance mode normal ( , sd = s ) s p
2 exp s > 0 s
2
2 .
1 ; 1 /
. /
2
2s
2 normal ( , var = v ) p
1
2v exp v > 0
. /
2
2v same same v same normal ( , prec =
) q
2 exp
> 0
. /
2
2 same same
1= same
4164
F
Chapter 52: The MCMC Procedure
Table 52.25
Pareto Distribution
PROC specification density parameter restriction range mean variance mode random number useful transformation pareto ( a , a C 1 b ) a b b a > 0; b > 0
2 Œb; 1 / ab a 1 if a > 1 b
2 a
.a 1/
2
.a 2/ if a > 2 b inverse CDF method with uniform .0; 1/ , and
D b u
1=a
F . / D 1 .b= / a
. Generate u is a draw from the Pareto distribution.
x D 1= is Beta( a
, 1) I { x < 1=b
}.
Table 52.26
Poisson Distribution
PROC specification density parameter restriction range poisson ( )
Š exp . /
2
0
( f
0; 1; : : : g f 0 g if > 0 if
D 0 mean variance mode
, if > 0 round
./
Table 52.27
PROC specification density parameter restriction range mean variance mode random number
Scaled Inverse
2
Distribution sichisq (
; s
2
)
.=2/
=2
.=2/ s
.=2 C 1/ e
> 0; s > 0 s
2
=.2 /
2 .0; 1 /
2 s
2 if > 2
2
2
. 2/
2
. 4/ s
4 if
> 4
C 2 s
2 scaled inverse igamma
2 is a special case of the inverse-gamma distribution:
.=2; scale
D
.s
2
/=2/ is a draw from the scaled inverse
2 distribution.
Standard Distributions
F
4165
Table 52.28
T Distribution
PROC specification density t ( , sd = s , ) parm striction rerange mean variance mode random number
.
.
2
/s
C 1 p
/
.1
s > 0
,
> 0
C
. /
2 s
2
/
2
.
1 if
> 1
;
1
/
2 s
2 if > 2 x
1 normal
.0; 1/; x
2 t
-distribution.
C 1
2 t ( , var = v , )
.
.
2
/
C 1 p
/ v
.1
C
. /
2 v
/ v > 0
,
> 0
C 1
2 t ( , prec =
.
.
2
C 1
2
/ p
/ p
.1
> 0
,
> 0
,
C
)
. /
2 same same same
2 v if > 2 same
2
.d /; and
D same
2
1 if > 2 same m C x
1 p d=x
2 is a draw from the
/
C 1
2
Table 52.29
Uniform Distribution
PROC specification density parameter restriction range mean variance mode random number uniform
( a
,
8
ˆ
ˆ
<
1 a b
1 b a if if
ˆ
ˆ
:
1 if b a > b b > a a
)
D b none
2 Œa; b a C b
2 j b a j
2
12 does not exist
Mersenne Twister ( Matsumoto and Kurita 1992 , 1994 ; Matsumoto and Nishimura 1998 )
4166
F
Chapter 52: The MCMC Procedure
Table 52.30
Wald Distribution
PROC specification density parameter restriction range mean variance mode random number wald ( q
, )
2
3 exp
> 0; > 0
2
.0;
1
/
. /
2
2
2
3
=
1 C
9
2
4
2
1=2
3
2 generate
0
2
.1/
. Let x
1
D C
2
2
0
2 and
If w x
2
D
D
1
2
=x
1
. Perform a Bernoulli trial,
, choose
D x
1 w
; otherwise, choose q
4
0
Bernoulli
D x
2
C
.
2
C x
1
2
0
/
.
Table 52.31
Weibull Distribution
PROC specification density parameter restriction range mean variance mode random number weibull
( , c
, exp c
) c c 1 c > 0; > 0
2 Œ; 1 / if c D 1 I .; 1 / otherwise
C .1
C 1=c/
2
Œ.1
C
2=c/
2
.1
C
1=c/
C .1
1=c/
1=c if c > 1 inverse CDF method with ate u uniform
.0; 1/
, and the Weibull distribution.
F . /
D
D 1
C exp
.
c
. Generln u/
1=c is a draw from
you can use the GENERAL and DGENERAL functions. The letter “D” stands for discrete. The new distributions have to be specified on the logarithm scale.
Suppose that you want to use the inverse-beta distribution: p.˛ j a; b/ D
.a
C b/
.a/ C .b/
˛
.a 1/
.1
C ˛/
.a
C b/
The following statements in PROC MCMC define the density on its log scale:
Using Density Functions in the Programming Statements
F
4167
a = 3; b = 5; const = lgamma(a + b) - lgamma(a) - lgamma(b); lp = const + (a - 1) * log(alpha) - (a + b) * log(1 + alpha); prior alpha ~ general(lp);
The symbol lp is the expression for the log of an inverse-beta (a = 3, b = 5). The function
general(lp)
assigns that distribution to alpha
. Note that the constant term, const
, can be omitted as the Markov simulation requires only the log of the density kernel.
When you use the GENERAL function in the dependent variable on the left of the tilde (
statement, you do not need to specify the
). The log-likelihood function takes the dependent variable into account; hence there is no need to explicitly state the dependent variable in the
statement. However, in the
statements, you need to explicitly state the parameter names and a tilde with the GENERAL and DGENERAL functions.
You can specify any distribution function by using the GENERAL and DGENERAL functions as long as they are programmable with SAS statements. When the function is used in the
statements, you must supply initial values. This can be done in either the
PARMS statement (“ PARMS Statement ”
on page 4139) or within the
and
ENDCNST statements (“ BEGINCNST/ENDCNST
It is important to remember that PROC MCMC does not verify that the GENERAL function you specify is a valid distribution—that is, an integrable density. You must use the function with caution.
Density Functions in PROC MCMC
PROC MCMC also has a number of internally defined log-density functions. The functions have the basic form of lpdf dist ( x , parm-list , < lower >, < upper >), where dist is the name of the distribution
(see
x and upper is the random variable, are boundary arguments. The lower and upper parm-list is the list of parameters, and lower arguments are optional but positional. With the exception of the Bernoulli and uniform distribution, you can specify limits on all distributions.
To set a lower bound on the normal density:
lpdfnorm(x, 0, 1, -2);
To set just an upper bound, specify a missing value for the lower bound argument:
lpdfnorm(x, 0, 1, ., 2);
Leaving both limits out gives you the unbounded density, and you can also specify both bounds:
lpdfnorm(x, 0, 1); lpdfnorm(x, 0, 1, -3, 4);
See the following table for a list of distributions and their corresponding lpdf functions.
4168
F
Chapter 52: The MCMC Procedure
Table 52.32
Logarithm of Density Functions in PROC MCMC
Distribution Name beta binary binomial
Cauchy
2 exponential
2 exponential gamma exponential exponential exponential inverse
2
Function Call
lpdfbeta(
x
,
a
,
b
,
< lower
>
,
< upper
>
); lpdfbern(
x
, lpdfbin(
x
,
p n
,
);
p
, lpdfcau(
x
,
loc
,
< lower
scale
,
<
>
,
< lower
> upper
,
<
>
); upper
>
); lpdfchisq(
x
, lpdfechisq(
x
,
df
,
df
,
< lower > ,
< lower
>
,
< upper > );
< upper
>
); lpdfegamma(
x
,
sp
,
scale
,
< lower
>
,
< upper
>
); lpdfeexpon(
x
,
scale
, < lower > , < upper > ); lpdfeichisq(
x
,
df
,
< lower
>
,
< upper
>
);
exponential inversegamma exponential scaled inverse
2 exponential gamma geometric inverse
2 inverse-gamma
Laplace logistic lognormal negative binomial normal
Pareto
Poisson scaled inverse
2
T uniform
Wald
Weibull
lpdfeigamma(
x
,
< upper
>
);
sp
,
scale
,
< lower
>
, lpdfesichisq(
x
,
df
,
scale
, < lower > ,
< upper
>
); lpdfexpon(
x
, lpdfgamma(
x
,
scale
,
sp
,
<
scale
lower
, <
>
, lower
<
> upper
, <
>
); upper > ); lpdfgeo(
x
,
p
, lpdfichisq(
x
, lpdfigamma(
x
,
< lower
>
,
df
,
sp
,
scale
,
<
< upper
< lower
>
,
<
>
); upper lower > , <
>
); upper > ); lpdfdexp(
x
, lpdflogis(
x
,
loc
,
loc
,
scale
,
scale
,
< lower
>
,
< lower
>
,
< upper
>
);
< upper
>
); lpdflnorm(
x
,
loc
,
sd
,
< lower
>
,
< upper
>
); lpdfnegbin(
x
, lpdfnorm(
x
,
mu n
,
,
p
,
sd
,
< lower
>
,
< lower
>
, lpdfpareto(
x
,
sp
,
scale
,
< upper
>
);
< upper
< lower
>
,
<
>
); upper
>
); lpdfpoi(
x
,
mean
lpdfsichisq(
x
,
,
df
,
< lower
>
,
scale
,
<
< upper lower
>
,
>
);
< upper
>
); lpdft(
x
,
mu
lpdfunif(
x
,
,
sd a
,
,
b df
);
, lpdfwald(
x
,
mean
,
< lower
scale
,
<
>
, lower
<
> upper
,
>
);
< upper
>
); lpdfwei(
x
,
< upper
>
);
loc
,
sp
,
scale
,
< lower
>
,
Standard Distributions, the logpdf Functions, and the lpdf dist Functions
Standard distributions listed in the section “ Standard Distributions ” on page 4155 are
names only, and they can only be used in the
statements to specify either a prior distribution or a conditional distribution of the data given parameters. They do not return any values, and you cannot use them in the programming statements.
Truncation and Censoring
F
4169
The LOGPDF functions are DATA step functions that compute the logarithm of various probability density (mass) functions. For example,
logpdf("beta", x, 2, 15)
density with parameters a = 2 and b = 15, evaluated at returns the log of a beta x
. All the LOGPDF functions are supported in PROC MCMC.
The lpdf dist functions are unique to PROC MCMC. They compute the logarithm of various probability density (mass) functions. The functions are the same as the LOGPDF functions when it comes to calculating the log density. For example,
lpdfbeta(x, 2, 15)
returns the same value as
logpdf("beta", x, 2, 15)
. The lpdf dist functions cover a greater class of probability density functions, and they take the optional but positional boundary arguments. There are no corresponding
lcdf
dist or
lsdf
dist functions in PROC MCMC. To work with the cumulative probability function or the survival functions, you need to use the LOGCDF and the LOGSDF DATA step functions.
Truncated Distributions
To specify a truncated distribution, you can use the LOWER= and/or UPPER= options. Almost all of the standard distributions, including the
and
DGENERAL functions, take these optional
truncation arguments. The exceptions are the binary and uniform distributions.
For example, you can specify the following:
prior alpha ~ normal(mean = 0, sd = 1, lower = 3, upper = 45);
or
parms beta; a = 3; b = 7; ll = (a + 1) * log(b / beta); prior beta ~ general(ll, upper = b + 17);
The preceding statements state that if beta is less than b+17
, the log of the prior density is calculated by the equation; otherwise, the log of the prior density is missing—the log of zero.
ll
, as
When the same distribution is applied to multiple parameters in a
statement, the LOWER= and UPPER= truncations apply to all parameters in that statement. For example, the following statements define a Poisson density for theta and gamma
:
parms theta gamma; lambda = 7; l1 = theta * log(lambda) - lgamma(1 + theta); l2 = gamma * log(lambda) - lgamma(1 + gamma); ll = l1 + l2; prior theta gamma ~ dgeneral(ll, lower = 1);
The LOWER=1 condition is applied to both theta and gamma
, meaning that for the assignment to ll to be meaningful, both theta and gamma have to be greater than 1. If either of the parameters is less than 1, the log of the joint prior density becomes a missing value.
4170
F
Chapter 52: The MCMC Procedure
With the exceptions of the normal distribution and the
and
functions, the
LOWER= and UPPER= options cannot be parameters or functions of parameters. The reason is that most of the truncated distributions are not normalized. Unnormalized densities do not lead to wrong MCMC answers as long as the bounds are constants. However if the bounds involve model parameters, then the normalizing constant, which is a function of these parameters, must be taken into account in the posterior. Without specifying the normalizing constant, inferences on these boundary parameters are incorrect.
It is not difficult to construct a truncated distribution with a normalizing constant. Any truncated distribution has the probability distribution: p. / p.
j a < < b/ D
F .a/ F .b/ where p.
/ p.
/ is the density function and
F .
is probability density function and
/ is the cumulative distribution function. In SAS functions,
F .
/ is cumulative distribution function. The following example shows how to construct a truncated gamma prior on theta
, with SHAPE = 3, SCALE = 2,
LOWER = a, and UPPER = b:
lp = logpdf('gamma', theta, 3, 2)
- log(cdf('gamma', a, 3, 2) - cdf('gamma', b, 3, 2)); prior theta ~ general(lp);
Note the difference from a naive definition of the density, without taking into account of the normalizing constant:
lp = logpdf('gamma', theta, 3, 2); prior theta ~ general(lp, lower=a, upper=b);
If a or b are parameters, you get very different results from the two formulations.
Censoring
There is no built-in mechanism in PROC MCMC that models censoring automatically. You need to construct the density function (using a combination of the LOGPDF, LOGCDF, and LOGSDF functions and IF-ELSE statements) for the censored data.
Suppose that you partition the data into four categories: uncensored (with observation censored (with observation xl
), right censored (with observation xr x
), left
), and interval censored (with observations xl and xr
). The likelihood is the normal with mean mu and standard deviation following statements construct the corresponding log likelihood for the observed data: s
. The
if uncensored then ll = logpdf('normal', x, mu, s); else if leftcensored then ll = logcdf('normal', xl, mu, s); else if rightcensored then ll = logsdf('normal', xr, mu, s); else /* this is the case of interval censored. */ ll = log(cdf('normal', xr, mu, s) - cdf('normal', xl, mu, s)); model general(ll);
See “ Example 52.9: Normal Regression with Interval Censoring ” on page 4288.
Multivariate Density Functions
F
4171
The DATA step has functions that compute the logarithm of the density of some multivariate distributions. You can use them in PROC MCMC. For a complete listing of multivariate functions, see SAS Language Reference: Dictionary .
Some commonly used multivariate functions in Bayesian analysis are as follows:
LOGMPDFNORMAL, the logarithm of the multivariate normal
LOGMPDFWISHART, the logarithm of the Wishart
LOGMPDFIWISHART, the logarithm of the inverted-Wishart
LOGMPDFDIR1, the logarithm of the Dirichlet distribution of Type I
LOGMPDFDIR2, the logarithm of the Dirichlet distribution of Type II
LOGMPDFMULTINOM, the logarithm of the multinomial
Other multivariate density functions include: LOGMPDFT ( t -distribution), LOGMPDFGAMMA
(gamma distribution), LOGMPDFBETA1 (beta of type I), and LOGMPDFBETA2 (beta of type II).
Density Function Definition
LOGMPDFNORMAL
Let x be an n -dimensional random vector with mean vector is pdf .x
I ;
†
/ D exp .
1
2
.x
/
T
† p
.2/ n j
† j
1
.x
// where j
† j is the determinant of the covariance matrix † .
The function has syntax: and covariance matrix † . The density y D
LOGMPDFNORMAL
.x
_ li st;
_ li st; cov
_ name/ I
W
ARNING
: you must set up the cov _ name covariance matrix before using the LOGMPDFNOR-
MAL function and free the memory after PROC MCMC exits. See the section “ Set Up the Covariance
Matrices and Free Memory ” on page 4173.
LOGMPDFWISHART and LOGMPDFIWISHART
The density function from the Wishart distribution is: pdf .x
I
; † /
D
1
C n
./ j † j
2 j x j n
2
1 exp
1
2 t r.
†
1 x/
4172
F
Chapter 52: The MCMC Procedure with > n , and the trace of a square matrix A is given by: t r.A/ D
X a i i i
C n
./ D 2 n
2
n
2
n
.z/ D
The density function from the inverse-Wishart distribution is: n.n
4
1/ n
Y
i D 1 pdf .x
I ; for > 2n , and
†
/ D
1
D n
./ j
† j n
2
1 j x j
2 exp
1
2 t r.
† x
1
/
D n
./
D
2
.
n
2
1/n
n n
2
1
If V I W n
.; † / then V
1
W n
.
n 1; †
1
/ z
The functions have syntax: y D
LOGMPDFWISHART
.
’ name
’
V
; ;
’ name
’
†
/ I and for the inverted Wishart: i
2
1 y D
LOGMPDFIWISHART
.
’ name
’
V
; ;
’ name
’
†
/ I
The three arguments are the multivariate matrix ’ name
’
V
, the degrees of freedom covariance matrix ’ name
’
† k
, and the
W
ARNING
: you must set up the cov _ name covariance matrix before using these functions and free
the memory after PROC MCMC exits. See the section “ Set Up the Covariance Matrices and Free
LOGMPDFDIR1 and LOGMPDFDIR2
The random variables u
1
:::u distribution with parameters k
, with a
1
:::a u i k C 1
> 0 and
P k i D 1 u i
< 1
, are said to have a Dirichlet Type I if their joint pdf is given by:
1 pdf
1
.u
1
; u
2
; :::; u k
; a
1
; a
2
; :::; a k C 1
/ D
.
P k C 1 i D 1
Q r C 1 i D 1 a i
.a
i
/
/
0 k
Y
@ u i a i i D 1
1 0
1
A @
1
1 k
X u i
A a k C 1 i D 1
The variables are said to have a Dirichlet type II distribution with parameters a
1
:::a k C 1 pdf is given by the following: if their joint pdf
2
.u
1
; u
2
; :::; u k
; a
1
; a
2
; :::; a k C 1
/ D
.
P k C 1 i D 1
Q r C 1 i D 1 a i
.a
i
/
/
0 k
Y
@ u a i i i D 1
1 0
1
A @
1
1
C k
X u i
A i D 1
P k C 1 i D 1 a i
The functions have syntax: y
D
LOGMPDFDIR1 .u
_ li st; a _ li st /
I and y D
LOGMPDFDIR2
.u
_ li st; a
_ li st / I
Multivariate Density Functions
F
4173
LOGMPDFMULTINOM
Let n
1
:::n k be random variables that denote the number of occurring of the events respectively occurring with probabilities joint distribution of n
1
; ::::::n k p
1
:::p k is the following:
. Let
P k i
D
1 p i
D 1 and let n D
P k i
D
1
E
1
; ::::E k n i
. Then the pdf .n
1
; n
2
; :::n k
; p
1
; p
2
; :::; p k
/ D nŠ k
Y p i n i
!
n i
Š i D 1
The function has syntax: y
D
LOGMPDFMULTINOM .n
_ li st; p _ li st /
I
Set Up the Covariance Matrices and Free Memory
For distributions that require symmetric positive definite matrices, such as the LOGMPDFNORMAL,
LOGMPDFWISHART and LOGMPDFIWISHART functions, you need to set up these matrices by using the following functions:
Use LOGMPDFSETSQ to set up a symmetric positive definite matrix from all its elements: rc
D
LOGMPDFSETSQ .name; num1; num2; ::::::::/
I rc is set to
0 when the numeric arguments describe a symmetric positive definite matrix, otherwise it is set to a nonzero value.
Use LOGMPDFSET to set up a symmetric positive definite matrix from its lower triangular elements: rc
D
LOGMPDFSET .name; num1; num2; ::::::::/
I
When the numeric arguments describe a symmetric positive definite matrix, the returned value rc is set to
0
. Otherwise, a nonzero value for rc is returned.
Use LOGMPFFREE to free the workspace previously allocated with either LOGMPDFSET or LOGMPDFSETSQ: rc
D
LOGMPDFFREE .< ::: < ’ name ’ >; ’ name2 ’ > :::/
I
When called without arguments, the LOGMPDFFREE frees all the symbols previously allocated by LOGMPDFSETSQ or LOGMPDFSET. Each freed symbol is reported back in the
SAS log.
The parameters used in these functions are defined as follows: name is a string containing the name of the work space that stores the matrix by the numeric parameters num1; :::
.
num1; ::: are numeric arguments that represent the elements of a symmetric positive definite matrix.
4174
F
Chapter 52: The MCMC Procedure
You would set up this matrix under the DATA step by using the following syntax: rc D
LOGMPDFSETSQ
.name;
11
;
12
;
21
;
22
/ I or the syntax: rc D
LOGMPDFSET
.name;
11
;
21
;
22
/ I
If the matrix is positive definite, the returned value rc is zero.
Table 52.33
Some Useful SAS Functions
SAS Function Definition
abs(x) airy(x) beta(x1, x2) call logistic(x) call softmax(x1,...,xn) call stdize(x1,...,xn) cdf cdf(’normal’, x, 0, 1) comb(x1, x2) constant(’.’) cos(x) css(x1, ..., xn) cv(x1, ..., xn) dairy(x) dimN(m)
(x1 eq x2) x1**x2 geomean(x1, ..., xn) difN(x) digamma(x1) erf(x) erfc(x) fact(x) floor(x) gamma(x) harmean(x1, ..., xn) ibessel(nu, x, kode) jbessel(nu, x) lagN(x)
j x j returns the value of the AIRY function.
R
1
0 z x1 1
.1
z/ x2 1 dz ln .x=.1
x// each element is replaced by exp
.x
j
/=
P exp
.x
j
/ standardize values cumulative distribution function standard normal cumulative distribution function x1Š x2Š.x1 x2/Š calculate commonly used constants cosine(x)
P i
.x
i x/
2 std(x) / mean(x) * 100 derivative of the AIRY function returns the numbers of elements in the Nth dim of array m returns 1 if x1 = x2; 0 otherwise x1 x2 exp log
.x
1
/ CC log
.x
n
/ n returns differences between the argument and its Nth lag
0
.x1/
.x1/ p
R x
0 exp. z
2
/dz
1 - erf(x) xŠ greatest integer
R
1
0 z x 1 n exp x
. 1/dz
1=x
1
C 1=x n modified Bessel function of order
Bessel function of order nu nu evaluated at x evaluated at x returns values from a queue
Some Useful SAS Functions
F
4175
Table 52.33
(continued)
SAS Function
largest(k, x1, ..., xn) lgamma(x) lgamma(x+1) log(x), logN(x) logbeta(x1, x2) logcdf logpdf logsdf max(x1, x2) mean(of x1-xn) median(of x1-xn) min(x1, x2) missing(x) mod(x1, x2) n(x1, ..., xn) nmiss(of y1-yn) quantile pdf perm(n, r) put round(x) rms(of x1-xn) sdf sign(x) sin(x) smallest(
s
, x1, ..., en ) sortn(of x1-xn) sqrt(x) std(x1, ..., xn) sum(of x:) trigamma(x) uss(of x1-xn)
Definition the k t h largest element ln ..x// ln .xŠ/ ln
.x/ lgamma( x
1
) + lgamma( x
2
) - lgamma( x
1
C log of a left cumulative distribution function x
2
) log of a probability density (mass) function log of a survival function returns x
1 if x
1
> x
2
; x
2 otherwise
P i x i
=n returns the median of nonmissing values returns x
1 if x
1
< x
2
; x
2 otherwise returns 1 if x is missing; 0 otherwise returns the remainder from x
1
=x
2 returns number of nonmissing values number of missing values computes the quantile from a specific distribution probability density (mass) functions nŠ
.n r/Š returns a value that uses a specified format rounds x q x
2
1
C x
2 n n survival function returns
1 if x < 0
;
0 if x D sine( x
) the s t h smallest component of
0
; x
1
1
; if x > 0
; x n p x standard deviation of x
1
; ; x n
(n-1 in denominator)
P i x i derivative of the DIGAMMA( x
) function uncorrected sum of squares
4176
F
Chapter 52: The MCMC Procedure
Here are examples of some commonly used transformations: logit log
mu = beta0 + beta1 * z1; call logistic(mu); w = beta0 + beta1 * z1; mu = exp(w);
probit
w = beta0 + beta1 * z1; mu = cdf(`normal', w, 0, 1);
cloglog
w = beta0 + beta1 * z1; mu = 1 - exp(-exp(w));
The MCMC procedure provides you with a number of CALL routines for performing simple matrix operations on declared arrays. With the exception of FILLMATRIX, IDENTITY, and ZEROMATRIX, the CALL routines listed in
do not support matrices or arrays that contain missing values.
Table 52.34
CALL Routine
Matrix Functions in PROC MCMC
Description
Performs an element-wise addition of two matrices or of a matrix and a scalar.
Calculates the Cholesky decomposition for a particular symmetric matrix.
Calculates the determinant of a specified matrix, which must be square.
Performs an element-wise multiplication of two matrices.
Replaces all of the element values of the input matrix with the specified value. You can use this routine with multidimensional numeric arrays.
Converts the input matrix to an identity matrix. Diagonal element values of the matrix are set to 1, and the rest of the values are set to 0.
Calculates a matrix that is the inverse of the input matrix. The input matrix must be a square, nonsingular matrix.
Calculates the matrix product of two input matrices.
Performs an element-wide subtraction of two matrices or of a matrix and a scalar.
Returns the transpose of a matrix.
Replaces all of the element values of the numeric input matrix with 0.
Matrix Functions in PROC MCMC
F
4177
ADDMATRIX CALL Routine
The ADDMATRIX CALL routine performs an element-wise addition of two matrices or of a matrix and a scalar.
The syntax of the ADDMATRIX CALL routine is
CALL ADDMATRIX
(X, Y, Z)
;
where
X specifies a scalar or an input matrix with dimensions m n
(that is, X [ m; n
])
Y
Z specifies a scalar or an input matrix with dimensions specifies an output matrix with dimensions m m n
(that is, Y [ m; n
]) n
(that is, Z [ m; n
]) such that
Z
D
X
C
Y
CHOL CALL Routine
The CHOL CALL routine calculates the Cholesky decomposition for a particular symmetric matrix.
The syntax of the CHOL CALL routine is
CALL CHOL
(X, Y < , validate > )
;
where
X specifies a symmetric positive-definite input matrix with dimensions m]) m m (that is,
X
[m,
Y is a variable that contains the Cholesky decomposition and specifies an output matrix with dimensions m m (that is,
Y
[ m; m ]) validate specifies an optional argument that can increase the processing speed by avoiding error checking:
If validate = 0 or is not specified, then the matrix X is checked for symmetry.
If validate
= 1, then the matrix
X is assumed to be symmetric.
such that
X
D
YY where Y is a lower triangular matrix with strictly positive diagonal entries and conjugate transpose of
Y
.
Y denotes the
Both input and output matrices must be square and have the same dimensions. If positive-definite,
Y is a lower triangle matrix. If
X
X is not symmetric positive-definite,
Y is symmetric is filled with missing values.
4178
F
Chapter 52: The MCMC Procedure
DET CALL Routine
The determinant, the product of the eigenvalues, is a single numeric value. If the determinant of a matrix is zero, then that matrix is singular (that is, it does not have an inverse). The routine performs an LU decomposition and collects the product of the diagonals.
The syntax of the DET CALL routine is
CALL DET
(X, a)
;
where
X specifies an input matrix with dimensions m a specifies the returned determinate value m
(that is, X [ m; m
]) such that a
D j
X j
ELEMMULT CALL Routine
The ELEMMULT CALL routine performs an element-wise multiplication of two matrices.
The syntax of the ELEMMULT CALL routine is
CALL ELEMMULT
(X, Y, Z)
;
where
X
Y specifies an input matrix with dimensions m specifies an input matrix with dimensions m
Z specifies an output matrix with dimensions m n (that is,
X
[ m; n ]) n
(that is, Y [ m; n
]) n (that is,
Z
[ m; n ])
FILLMATRIX CALL Routine
The FILLMATRIX CALL routine replaces all of the element values of the input matrix with the specified value. You can use the FILLMATRIX CALL routine with multidimensional numeric arrays.
The syntax of the FILLMATRIX CALL routine is
CALL FILLMATRIX
(X, Y)
;
where
X specifies an input numeric matrix
Y specifies the numeric value that is used to fill the matrix
Matrix Functions in PROC MCMC
F
4179
IDENTITY CALL Routine
The IDENTITY CALL routine converts the input matrix to an identity matrix. Diagonal element values of the matrix are set to 1, and the rest of the values are set to 0.
The syntax of the IDENTITY CALL routine is
CALL IDENTITY
(X)
;
where
X specifies an input matrix with dimensions m m (that is,
X
[ m; m ])
INV CALL Routine
The INV CALL routine calculates a matrix that is the inverse of the input matrix. The input matrix must be a square, nonsingular matrix.
The syntax of the INV CALL routine is
CALL INV
(X, Y)
;
where
X specifies an input matrix with dimensions m
Y specifies an output matrix with dimensions m m (that is,
X
[ m; m ]) m
(that is,
Y
[ m; m
])
MULT CALL Routine
The MULT CALL routine calculates the matrix product of two input matrices.
The syntax of the MULT CALL routine is
CALL MULT
(X, Y, Z)
;
where
X specifies an input matrix with dimensions m
Y specifies an input matrix with dimensions n n
(that is, X [ m; n
]) p
(that is, Y [ n; p
])
Z specifies an output matrix with dimensions m p (that is,
Z
[ m; p ])
The number of columns for the first input matrix must be the same as the number of rows for the second matrix. The calculated matrix is the last argument.
4180
F
Chapter 52: The MCMC Procedure
SUBTRACTMATRIX CALL Routine
The SUBTRACTMATRIX CALL routine performs an element-wide subtraction of two matrices or of a matrix and a scalar.
The syntax of the SUBTRACTMATRIX CALL routine is
CALL SUBTRACTMATRIX
(X, Y, Z)
;
where
X specifies a scalar or an input matrix with dimensions m n (that is,
X
[ m; n ])
Y specifies a scalar or an input matrix with dimensions m n (that is,
Y
[ m; n ])
Z specifies an output matrix with dimensions m n (that is,
Z
[ m; n ]) such that
Z
D
X Y
TRANSPOSE CALL Routine
The TRANSPOSE CALL routine returns the transpose of a matrix.
The syntax of the TRANSPOSE CALL routine is
CALL TRANSPOSE
(X, Y)
;
where
X specifies an input matrix with dimensions m
Y specifies an output matrix with dimensions n n
(that is, X [ m; n
]) m (that is,
Y
[ n; m ])
ZEROMATRIX CALL Routine
The ZEROMATRIX CALL routine replaces all of the element values of the numeric input matrix with 0. You can use the ZEROMATRIX CALL routine with multidimensional numeric arrays.
The syntax of the ZEROMATRIX CALL routine is
CALL ZEROMATRIX
(X)
;
where
X specifies a numeric input matrix.
Modeling Joint Likelihood
F
4181
PROC MCMC assumes that the input observations are independent and that the joint log likelihood is the sum of individual log-likelihood functions. You specify the log likelihood of one observation in the
statement. PROC MCMC evaluates that function for each observation in the data set and cumulatively sums them up. If observations are not independent of each other, this summation produces the incorrect log likelihood.
There are two ways to model dependent data. You can either use the DATA step LAG function or use the PROC option
JOINTMODEL . The LAG function returns values of a variable from a queue. As
PROC MCMC steps through the data set, the LAG function queues each data set variable, and you have access to the current value as well as to all previous values of any variable. If the log likelihood for observation x i depends only on observations 1 to i in the data set, you can use this SAS function to construct the log-likelihood function for each observation. Note that the LAG function enables you to access observations from different rows, but the log-likelihood function in the
statement
must be generic enough that it applies to all observations. See “ Example 52.8: Cox Models ” on
page 4271 for how to use this LAG function.
A second option is to create arrays, store all relevant variables in the arrays, and construct the joint log likelihood for the entire data set instead of for each observation. Following is a simple example that illustrates the usage of this option. For a more realistic example that models dependent data, see
“ Example 52.8: Cox Models ” on page 4271.
/* allocate the sample size. */ data exi; call streaminit(17); do ind = 1 to 100; y = rand("normal", 2.3, 1); output; end; run;
The log-likelihood function for each observation is as follows: log
.f .y
i j ; // D log
..y
i
I ; var
D
2
//
The joint log-likelihood function is as follows: log
.f .
y j ; // D
X log
..y
i
I ; var i
D
2
//
The following statements fit a simple model with an unknown mean ( mu
) in PROC MCMC, with the variance in the likelihood assumed known. The
statement indicates a normal likelihood for each observation y
.
proc mcmc data=exi seed=7 outpost=p1; parm mu; prior mu ~ normal(0, sd=10); model y ~ normal(mu, sd=1); run;
4182
F
Chapter 52: The MCMC Procedure
The following statements show how you can specify the log-likelihood function for the entire data set:
data a; run; proc mcmc data=a seed=7 outpost=p2 jointmodel; array data[1] / nosymbols; begincnst; rc = read_array("exi", data, "y"); n = dim(data, 1); endcnst; parm mu; prior mu ~ normal(0, sd=10); ll = 0; do i = 1 to n; ll = ll + lpdfnorm(data[i], mu, 1); end; model general(ll); run;
The
option indicates that the function used in the
statement calculates the log likelihood for the entire data set, rather than just for one observation. Given this option, the procedure no longer steps through the input data during the simulation. Consequently, you can no longer use any data set variables to construct the log-likelihood function. Instead, you store the data set in arrays and use arrays instead of data set variables to calculate the log likelihood.
The
statement allocates a temporary array ( data
). The READ_ARRAY function selects the y variable from the exi data set and stores it in the data
array. See the section “ READ_ARRAY
Function ” on page 4134. In the programming statements, you use a DO loop to construct the joint
log likelihood. The expression ll likelihood for all data.
in the
function now takes the value of the joint log
You can run the following statements to see that two PROC MCMC runs produce identical results.
proc compare data=p1 compare=p2; var mu; run;
By default, PROC MCMC generates three plots: the trace plot, the autocorrelation plot and the kernel density plot. Unless you had requested the display of ODS Graphics (ODS GRAPHICS ON) before calling the procedure, it is hard to generate the same graph afterwards. Directly using the template (Stat.MCMC.Graphics.TraceAutocorrDensity) is not feasible. To regenerate the same graph with a Markov chain, you need to define a template and use PROC SGRENDER to create the graph.
See the SGRENDER procedure in the SAS/GRAPH: Statistical Graphics Procedures Guide . The
following PROC TEMPLATE (see Chapter 21, “ Statistical Graphics Using ODS ”) statements define
Regenerating Diagnostics Plots
F
4183 a new graph template
mygraphs.mcmc
:
proc template; define statgraph mygraphs.mcmc; dynamic _sim _parm;
BeginGraph; layout gridded /rows=2 columns=1 rowgutter=5; seriesplot x=_sim y=_parm; layout gridded /rows=1 columns=2 columngutter=15; layout overlay / yaxisopts=(linearopts=(viewmin=-1 viewmax=1 tickvaluelist=(-1 -0.5 0 0.5 1)) label="Autocorrelation") xaxisopts=(linearopts=(integer=true) label="Lag" offsetmin=.015); needleplot x=eval(lags(_parm,Max=50)) y=eval(acf(_parm, NLags=50)); endlayout; layout overlay / xaxisopts=(label=_parm) yaxisopts=(label="Density"); densityplot _parm /kernel(); endlayout; endlayout; endlayout;
EndGraph; end;
The DEFINE STATGRAPH statement tells PROC TEMPLATE that you are defining a new graph template (instead of a table or style template). The template is named
mygraphs.mcmc
. There are two dynamic variables: variable
_parm
_sim and
_parm
. The variable
_sim is the iteration number and the is the variable in the data set that stores the posterior sample. All STATGRAPH template definitions must start with a BEGINGRAPH statement and conclude with a ENDGRAPH statement. The first LAYOUT GRIDDED statement assembles the results of nested STATGRAPH statements into a grid, with two rows and 1 column. The trace plot (SERIESPLOT) is shown in the first row of the graph. The second LAYOUT GRIDDED statement divides the second row of the graph into two graphs: one an autocorrelation plot (NEEDLEPLOT) and the other a kernel density plot (DENSITYPLOT). For details of other controls, such as the labels, line types, see Chapter 21,
“ Statistical Graphics Using ODS .”
A simple regression example, with three parameters, is used here. For an explanation of the regression
model and the data involved, see “ Simple Linear Regression ” on page 4104. The following statements
generate a SAS data set and fit a regression model:
4184
F
Chapter 52: The MCMC Procedure
title 'Simple Linear Regression'; data Class; input Name $ Height Weight @@; datalines;
Alfred 69.0 112.5
Carol
Jane
John
Louise
Robert
62.8 102.5
59.8
59.0
56.3
84.5
99.5
77.0
64.8 128.0
William 66.5 112.0
Alice
Henry
Janet
Joyce
Mary
56.5
51.3
84.0
63.5 102.5
62.5 112.5
50.5
66.5 112.0
Ronald 67.0 133.0
;
Barbara 65.3
James 57.3
Jeffrey 62.5
Judy 64.3
Philip
Thomas
98.0
83.0
84.0
90.0
72.0 150.0
57.5
85.0
proc mcmc data=class nmc=50000 thin=5 outpost=classout seed=246810; ods select none; parms beta0 0 beta1 0; parms sigma2 1; prior beta0 beta1 ~ normal(0, var = 1e6); prior sigma2 ~ igamma(3/10, scale = 10/3); mu = beta0 + beta1*height; model weight ~ normal(mu, var = sigma2); run; ods select all;
The output data set classout contains iteration number (
Iteration
) and posterior draws for beta0
, beta1
, and sigma2
. It also stores the log of the prior density (
LogPrior
), log of the likelihood (
LogLike
), and the log of the posterior density (
LogPost
). If you want to examine the
LogPost variable, you can use the following statements to generate the graphs:
proc sgrender data=classout template=mygraphs.mcmc; dynamic _sim='iteration' _parm='logpost'; run;
The SGRENDER procedure takes the classout data set and applies the template MY-
GRAPHS.MCMC that was defined previously. The DYNAMIC statement needs two arguments, iteration and logpost
. The resulting graph is shown in
Posterior Predictive Distribution
F
4185
Figure 52.11
Regenerate Diagnostics Plots for Log of the Posterior Density
The posterior predictive distribution p.
y pred j y / D
Z p.
y pred j /p.
j y /d can often be used to check whether the model is consistent with data. For more information about using predictive distribution as a model checking tool, see
Gelman et al. 2004 , Chapter 6 and the
bibliography in that chapter. The idea is to generate replicate data from p.
y pred j y / —call them y i pred
, for i D 1; ; M
, where
M is the total number of replicates—and compare them to the observed data to see whether there are any large and systematic differences. Large discrepancies suggest a possible model misfit. One way to compare the replicate data to the observed data is to first summarize the data to some test quantities, such as the mean, standard deviation, order statistics, and so on. Then compute the tail-area probabilities of the test statistics (based on the observed data) with respect to the estimated posterior predictive distribution that uses the M replicate y pred samples.
Let
T .
/ denote the function of the test quantity,
T .
y i pred
/ the test quantity that uses the i
T .
y / the test quantity that uses the observed data, and th replicate data from the posterior predictive distribution.
You calculate the tail-area probability by using the following formula:
Pr
.T .
y pred
/ > T .
y
/ j /
4186
F
Chapter 52: The MCMC Procedure
The following example shows how you can use PROC MCMC to estimate this probability.
An Example for the Posterior Predictive Distribution
This example uses a normal mixed model to analyze the effects of coaching programs for the scholastic aptitude test (SAT) in eight high schools. For the original analysis of the data, see
( 1981 ). The presentation here follows the analysis and posterior predictive check presented in
( 2004 ). The data are as follows:
title 'An Example for the Posterior Predictive Distribution'; data SAT; input effect se @@; ind=_n_; datalines;
28.39 14.9
7.94 10.2 -2.75 16.3
6.82 11.0 -0.64
9.4
0.63 11.4
18.01 10.4 12.16 17.6
;
The variable effect is the reported test score difference between coached and uncoached students in eight schools. The variable se is the corresponding estimated standard error for each school. In a normal mixed effect model, the variable effect is assumed to be normally distributed: effect i normal .
i
; se
2
/ for i
D
1; ; 8
The parameter i has a normal prior with hyperparameters .m; v/ : i normal
.m; var = v/
The hyperprior distribution on m is a uniform prior on the real axis, and the hyperprior distribution on v is a uniform prior from 0 to infinity.
The following statements fit a normal mixed model and use the
draws from the posterior predictive distribution.
statement to generate
ods listing close; proc mcmc data=SAT outpost=out nmc=50000 thin=10 seed=12; array theta[8]; parms theta: 0; parms m 0; parms v 1; hyper m ~ general(0); hyper v ~ general(1,lower=0); prior theta: ~ normal(m,var=v); mu = theta[ind]; model effect ~ normal(mu,sd=se); preddist outpred=pout nsim=5000; run; ods listing;
Posterior Predictive Distribution
F
4187
The ODS LISTING CLOSE statement disables the listing output because you are primarily interested in the samples of the predictive distribution. The
statements specify the Bayesian model of interest. The
statement generates samples from the posterior preditive distribution and stores the samples in the pout data set. The predictive variables are named effect_1
, , input data set effect_8
. When no
SAT
are used in the prediction. The option is specified, the covariates in the original
option specifies the number of predictive simulation iterations.
The following statements use the pout data set to calculate the four test quantities of interest: the average ( mean
), the sample standard deviation ( sd
), the maximum effect ( max
), and the minimum effect ( min
). The output is stored in the pred data set.
data pred; set pout; mean = mean(of effect:); sd = std(of effect:); max = max(of effect:); min = min(of effect:); run;
The following statements compute the corresponding test statistics, the mean, standard deviation, and the minimum and maximum statistics on the real data and store them in macro variables. You then calculate the tail-area probabilities by counting the number of samples in the data set pred that are greater than the observed test statistics based on the real data.
proc means data=SAT noprint; var effect; output out=stat mean=mean max=max min=min stddev=sd; run; data _null_; set stat; call symputx('mean',mean); call symputx('sd',sd); call symputx('min',min); call symputx('max',max); run; data _null_; set pred end=eof nobs=nobs; ctmean + (mean>&mean); ctmin + (min>&min); ctmax + (max>&max); ctsd + (sd>&sd); if eof then do; pmean = ctmean/nobs; call symputx('pmean',pmean); pmin = ctmin/nobs; call symputx('pmin',pmin); pmax = ctmax/nobs; call symputx('pmax',pmax); psd = ctsd/nobs; call symputx('psd',psd); end; run;
4188
F
Chapter 52: The MCMC Procedure
You can plot histograms of each test quantity to visualize the posterior predictive distributions. In addition, you can see where the estimated p -values fall on these densities.
shows the histograms. To put all four histograms on the same panel, you need to use PROC TEMPLATE to
define a new graph template. (See Chapter 21, “ Statistical Graphics Using ODS .”) The following
statements define the template
twobytwo
:
proc template; define statgraph twobytwo; begingraph; layout lattice / rows=2 columns=2; layout overlay / yaxisopts=(display=none) xaxisopts=(label="mean"); layout gridded / columns=2 border=false autoalign=(topleft topright); entry halign=right "p-value ="; entry halign=left eval(strip(put(&pmean, 12.2))); endlayout; histogram mean / binaxis=false; lineparm x=&mean y=0 slope=. / lineattrs=(color=red thickness=5); endlayout; layout overlay / yaxisopts=(display=none) xaxisopts=(label="sd"); layout gridded / columns=2 border=false autoalign=(topleft topright); entry halign=right "p-value ="; entry halign=left eval(strip(put(&psd, 12.2))); endlayout; histogram sd / binaxis=false; lineparm x=&sd y=0 slope=. / lineattrs=(color=red thickness=5); endlayout; layout overlay / yaxisopts=(display=none) xaxisopts=(label="max"); layout gridded / columns=2 border=false autoalign=(topleft topright); entry halign=right "p-value ="; entry halign=left eval(strip(put(&pmax, 12.2))); endlayout; histogram max / binaxis=false; lineparm x=&max y=0 slope=. / lineattrs=(color=red thickness=5); endlayout; layout overlay / yaxisopts=(display=none) xaxisopts=(label="min"); layout gridded / columns=2 border=false autoalign=(topleft topright); entry halign=right "p-value ="; entry halign=left eval(strip(put(&pmin, 12.2))); endlayout; histogram min / binaxis=false; lineparm x=&min y=0 slope=. / lineattrs=(color=red thickness=5);
Posterior Predictive Distribution
F
4189
endlayout; endlayout; endgraph; end; run;
You call PROC SGRENDER to create the graph, which is shown in
DER procedure in the SAS/GRAPH: Statistical Graphics Procedures Guide .) There are no extreme p -values observed; this supports the notion that the predicted results are similar to the actual observations and that the model fits the data.
proc sgrender data=pred template=twobytwo; run;
Figure 52.12
Posterior Predictive Distribution Check for the SAT example
Note that the posterior predictive distribution is not the same as the prior predictive distribution. The prior predictive distribution is p.
y / , which is also known as the marginal distribution of the data.
The prior predictive distribution is an integral of the likelihood function with respect to the prior distribution p.
y pred
/
D
Z p.
y pred j
/p. /d and the distribution is not conditional on observed data.
4190
F
Chapter 52: The MCMC Procedure
By default, PROC MCMC discards all observations that have missing values before carrying out the posterior sampling. This corresponds to the option
MISSING=CC , where CC stands for complete
cases. PROC MCMC does not automatically augment missing data. However, you can choose to model the missing values by using
MISSING=AC . Given this option, PROC MCMC does not discard
any missing values. It is up to you to specify how the missing values are handled in the program.
You can choose to model the missing values as parameters (a fully Bayesian approach) or assign specific values to them (multiple imputation). In general, however, the handling of missing values largely depends on the assumptions you have about the missing mechanism, which is beyond the scope of this chapter.
When performing a Markov chain Monte Carlo simulation, you must calculate a proposed jump and an objective function (usually a posterior density). These calculations might lead to arithmetic exceptions and overflows. A typical cause of these problems is parameters with widely varying scales. If the posterior variances of your parameters vary by more than a few orders of magnitude, the numerical stability of the optimization problem can be severely reduced and can result in computational difficulties. A simple remedy is to rescale all the parameters so that their posterior variances are all approximately equal. Changing the SCALE= option might help if the scale of your parameters is much different than one. Another source of numerical instability is highly correlated parameters. Often a model can be reparameterized to reduce the posterior correlations between parameters.
If parameter rescaling does not help, consider the following actions: provide different initial values or try a different seed value use boundary constraints to avoid the region where overflows might happen change the algorithm (specified in programming statements) that computes the objective function
Problems Evaluating Code for Objective Function
The initial values must define a point for which the programming statements can be evaluated.
However, during simulation, the algorithm might iterate to a point where the objective function cannot be evaluated. If you program your own likelihood, priors, and hyperpriors by using SAS statements and the
function in the
statements, you can specify that an expression cannot be evaluated by setting the value you pass back through the
function to missing. This tells the PROC MCMC that the proposed set of parameters is invalid, and the proposal will not be accepted. If you use the shorthand notation that the
statements provide, this error checking is done for you automatically.
Floating Point Errors and Overflows
F
4191
Long Run Times
PROC MCMC can take a long time to run for problems with complex models, many parameters, or large input data sets. Although the techniques used by PROC MCMC are some of the best available, they are not guaranteed to converge or proceed quickly for all problems. Ill-posed or misspecified models can cause the algorithms to use more extensive calculations designed to achieve convergence, and this can result in longer run times. You should make sure that your model is specified correctly, that your parameters are scaled to the same order of magnitude, and that your data reasonably match the model that you are specifying.
To speed general computations, you should check over your programming statements to minimize the number of unnecessary operations. For example, you can use the proportional kernel in the priors or the likelihood and not add constants in the densities. You can also use the
and
to reduce unnecessary computations on constants, and the
and
statements to reduce observation-level calculations.
Reducing the number of blocks (the number of the
statements) can speed up the sampling process. A single-block program is approximately three times faster than a three-block program for the same number of iterations. On the other hand, you do not want to put too many parameters in a single block, because blocks with large size tend not to produce well-mixed Markov chains.
Slow or No Convergence
There are a number of things to consider if the simulator is slow or fails to converge:
Change the number of Monte Carlo iterations ( NMC= ), or the number of burn-in iterations
( NBI= ), or both. Perhaps the chain just needs to run a little longer. Note that after the
simulation, you can always use the DATA step or the FIRSTOBS data set option to throw away initial observations where the algorithm has not yet burned in, so it is not always necessary to set
to a large value.
Increase the number of tuning. The proposal tuning can often work better in large models
(models that have more parameters) with larger values of
NTU= . The idea of tuning is to find
a proposal distribution that is a good approximation to the posterior distribution. Sometimes
500 iterations per tuning phase (the default) is not sufficient to find a good approximating covariance.
Change the initial values to more feasible starting values. Sometimes the proposal tuning starts badly if the initial values are too far away from the main mass of the posterior density, and it might not be able to recover.
Use the
option to start the Markov chain at better starting values. With the
option, PROC MCMC optimizes the object function and uses the posterior mode as the starting value of the Markov chain. In addition, a quadrature approximation to the posterior mode is used as the proposal covariance matrix. This option works well in many cases and can improve the mixing of the chain and shorten the tuning and burn-in time.
4192
F
Chapter 52: The MCMC Procedure
Change the blocking by using the
statements. Sometimes poor mixing and slow convergence can be attributed to highly correlated parameters being in different parameter blocks.
Modify the target acceptance rate. A target acceptance rate of about 25% works well for many multi-parameter problems, but if the mixing is slow, a lower target acceptance rate might be better.
Change the initial scaling or the
option to possibly help the proposal tuning.
Consider using a different proposal distribution. If from a trace plot you see that a chain traverses to the tail area and sometimes takes quite a few simulations before it comes back, you can consider using a t
-proposal distribution. You can do this by either using the PROC option
or using a
statement option
T
.
Transform parameters and sample on a different scale. For example, if a parameter has a gamma distribution, sample on the logarithm scale instead. A parameter a that has a gamma distribution is equivalent to log .a/ that has an egamma distribution, with the same distribution specification. For example, the following two formulations are equivalent:
parm a; prior a ~ gamma(shape = 0.001, iscale = 0.001);
and
parm la; prior la ~ egamma(shape = 0.001, iscale = 0.001); a = exp(la);
See “ Example 52.4: Nonlinear Poisson Regression Models
logit transformation on parameters that have uniform .0; 1/ priors. This prior is often used on probability parameters. The logit transformation is as follows: q D log
.
p
1 p
/
. The distribution on q is the Jacobian of the transformation: exp . q/.1
C exp . q//
2
. Again, the following two formulations are equivalent:
parm p; prior p ~ uniform(0, 1);
and
parm q; lp = -q - 2 * log(1 + exp(-q)); prior q ~ general(lp); p = 1/(1+exp(-q));
Precision of Solution
In some applications, PROC MCMC might produce parameter values that are not precise enough.
Usually, this means that there were not enough iterations in the simulation. At best, the precision of
Handling Error Messages
F
4193
MCMC estimates increases with the square of the simulation sample size. Autocorrelation in the parameter values deflate the precision of the estimates. For more information about autocorrelations
in Markov chains, see the section “ Autocorrelations ” on page 168.
PROC MCMC does not have a debugger. This section covers a few ways to debug and resolve error messages.
Using the PUT Statement
Adding the PUT statement often helps to find errors in a program. The following program produces an error:
data a; run; proc mcmc data=a seed=1; parms sigma lt w; beginnodata; prior sigma ~ unif(0.001,100); s2 = sigma*sigma; prior lt ~ gamma(shape=1, iscale=0.001); t = exp(lt); c = t/s2; d = 1/(s2); prior w ~ gamma(shape=c, iscale=d); endnodata; model general(0); run;
ERROR: PROC MCMC is unable to generate an initial value for the parameter w.
The first parameter in the prior distribution is missing.
To find out why the shape parameter c is missing, you can add the put statement and examine all the calculations that lead up to the assignment of c
:
proc mcmc data=a seed=1; parms sigma lt w; beginnodata; prior sigma ~ unif(0.001,100); s2 = sigma*sigma; prior lt ~ gamma(shape=1, iscale=0.001);
4194
F
Chapter 52: The MCMC Procedure
t = exp(lt); c = t/s2; d = 1/(s2); put c= t= s2= lt=; /* display the values of these symbols. */ prior w ~ gamma(shape=c, iscale=d); endnodata; model general(0); run;
In the log file, you see the following:
c=. t=. s2=. lt=.
c=. t=. s2=2500.0500003 lt=1000 c=. t=. s2=2500.0500003 lt=1000
ERROR: PROC MCMC is unable to generate an initial value for the parameter w.
The first parameter in the prior distribution is missing.
You can ignore the first few lines. They are the results of initial set up by PROC MCMC. The last line is important. The variable c is missing because t is the exponential of a very large number,
1000, in lt
. The value 1000 is assigned to lt by PROC MCMC because none was given. The gamma
prior with shape of 1 and inverse scale of 0.001 has mode 0 (see “ Standard Distributions ” on
page 4155 for more details). PROC MCMC avoids starting the Markov chain at the boundary of the support of the distribution, and it uses the mean value here instead. The mean of the gamma prior is 1000, hence the problem. You can change how the initial value is generated by using the
PROC statement
INIT=RANDOM . Do not forget to take out the
put statement once you identify the problem. Otherwise, you will see a voluminous output in the log file.
You can use the
statement to narrow down possible errors in the prior distribution specification. With multiple
statements in a program, you might see the following error message if one of the prior distributions is not specified correctly:
ERROR: The initial prior parameter specifications must yield log of positive prior density values.
This message is displayed when PROC MCMC detects an error in the prior distribution calculation but cannot pinpoint the specific parameter at fault. It is frequently, although not necessarily, associated with parameters that have
or
distributions. If you have a complicated model with many
statements, finding the parameter at fault can be time consuming. One way is to change a subset of the
statements to
statements. The two statements are treated the same in PROC MCMC and the simulation is not affected, but you get a different message if the hyperprior distributions are calculated incorrectly:
ERROR: The initial hyperprior parameter specifications must yield log of positive hyperprior density values.
Computational Resources
F
4195
This message can help you identify more easily which distributions are producing the error, and you can then use the PUT statement to further investigate.
It is not possible to estimate how long it will take for a general Markov chain to converge to its stationary distribution. It takes a skilled and thoughtful analysis of the chain to decide if it has converged to the target distribution and if the chain is mixing rapidly enough. It is easier, however, to estimate how long a particular simulation might take. The running time of a program is roughly linear to the following factors: the number of samples in the input data set ( nsamples
), the number of simulations ( nsim
), the number of blocks in the program ( nblocks
), and the speed of your computer.
For an analysis that uses a data set of size nsamples
, a simulation length of nsim
, and a block design of nblocks
, PROC MCMC evaluates the log-likelihood function the following number of times, excluding the tuning phase: nsamples nsim nblocks
The faster your computer evaluates a single log-likelihood function, the faster this program runs.
Suppose that you have nsamples equal to 200, nsim equal to 55,000, and
MCMC evaluates the log-likelihood function roughly a total number of nblocks
3:3 equal to 3. PROC
10
7 times. If your computer can evaluate the log likelihood, for one observation,
10
6 times per second, this program will take approximately a half a minute to run. If you want to increase the number of simulations five-fold, the run time will approximately increase five-fold as well.
Of course, larger problems take longer than shorter ones, and if your model is amenable to frequentist treatment, then one of the other SAS procedures might be more suitable. With “regular” likelihoods and a lot of data, the results of standard frequentist analysis are often asymptotically equivalent to a Bayesian approach. If PROC MCMC requires too much CPU time, then perhaps another tool in
SAS/STAT would be suitable.
This section describes the displayed output from PROC MCMC. For a quick reference of all ODS
table names, see the section “ ODS Table Names ” on page 4200. ODS tables are arranged under four
groups, listed in the following sections: “ Sampling Related ODS Tables
” on page 4197, “ Convergence Diagnostics Related ODS Tables ” on
page 4198, and “ Optimization Related ODS Tables ” on page 4199.
Sampling Related ODS Tables
Burn-In History
The “Burn-In History” table (ODS table name
BurnInHistory
) shows the scales and acceptance rates for each parameter block in the burn-in phase. The table is displayed by default.
4196
F
Chapter 52: The MCMC Procedure
Number of Observation Table
The “NObs” table (ODS table name
NOBS
) shows the number of observations that is in the data set and the number of observations that is used in the analysis. By default, observations with missing
table is displayed by default.
Parameters
The “Parameters” table (ODS table name
Parameters
) shows the name of each parameter, the block number of each parameter, the sampling method used for the block, the initial values, and the prior or hyperprior distributions. This table is displayed by default.
Parameters Initial Value Table
The “Parameters Initial” table (ODS table name
ParametersInit
) shows the value of each parameter after the tuning phase. This table is not displayed by default and can be requested by specifying the option
Posterior Samples
The “Posterior Samples” table (ODS table name
PosteriorSample
) stores posterior draws of all parameters. It is not printed by PROC MCMC. You can create an ODS output data set of the chain by specifying the following:
ODS OUTPUT PosteriorSample = SAS-data-set;
Sampling History
The “Sampling History” table (ODS table name
SamplingHistory
) shows the scales and acceptance rates for each parameter block in the main sampling phase. The table is displayed by default.
Tuning Covariance
The “Tuning Covariance” table (ODS table name
TuneCov
) shows the proposal covariance matrices for each parameter block after the tuning phase. The table is not displayed by default and can be requested by specifying the option
INIT=PINIT . For more details about proposal tuning, see the
section “ Tuning the Proposal Distribution ” on page 4150.
Tuning History
The “Tuning History” table (ODS table name
TuningHistory
) shows the number of tuning phases used in establishing the proposal distribution. The table also displays the scales and acceptance rates for each parameter block at each of the tuning phases. For more information about the self-adapting
proposal tuning algorithm used by PROC MCMC, see the section “ Tuning the Proposal Distribution ”
on page 4150. The table is displayed by default.
Displayed Output
F
4197
Tuning Probability Vector
The “Tuning Probability” table (ODS table name
TuneP
) shows the proposal probability vector for each discrete parameter block (when the option
DISCRETE=GEO is specified and the geometric proposal distribution is used for discrete parameters) after the tuning phase. The table is not displayed by default and can be requested by specifying the option
INIT=PINIT . For more information about
proposal tuning, see the section “ Tuning the Proposal Distribution ” on page 4150.
Posterior Statistics Related ODS Tables
PROC MCMC calculates some essential posterior statistics and outputs them to a number of ODS tables that you can request and save individually. For details of the calculations, see the section
“ Summary Statistics ” on page 169.
Summary Statistics
The “Posterior Summaries” table (ODS table name
PostSummaries
) contains basic statistics for each parameter. The table lists the number of posterior samples, the posterior mean and standard deviation estimates, and the percentile estimates. This table is displayed by default.
Correlation Matrix
The “Posterior Correlation Matrix” table (ODS table name
Corr
) contains the posterior correlation of model parameters. The table is not displayed by default and can be requested by specifying the option
STATS=CORR
.
Covariance Matrix
The “Posterior Covariance Matrix” table (ODS table name
Cov
) contains the posterior covariance of model parameters. The table is not displayed by default and can be requested by specifying the option
Deviance Information Criterion
The “Deviance Information Criterion” table (ODS table name
DIC
) contains the DIC of the model.
The table is not displayed by default and can be requested by specifying the option
of the calculations, see the section “ Deviance Information Criterion (DIC) ” on page 171.
Interval Statistics
The “Posterior Intervals” table (ODS table name
PostIntervals
) contains two the equal-tail and highest posterior density (HPD) interval estimates for each parameter. The default ˛ value is 0:05 , and you can change it to other levels by using the
option. This table is displayed by default.
4198
F
Chapter 52: The MCMC Procedure
Convergence Diagnostics Related ODS Tables
PROC MCMC has convergence diagnostic tests that check for Markov chain convergence. The procedure produces a number of ODS tables that you can request and save individually. For details
in calculation, see the section “ Statistical Diagnostic Tests ” on page 159.
Autocorrelation
The “Autocorrelations” table (ODS table name
AUTOCORR
) contains the first order autocorrelations of the posterior samples for each parameter. The “Parameter” column states the name of the parameter.
By default, PROC MCMC displays lag 1, 5, 10, and 50 estimates of the autocorrelations. You can request different autocorrelations by using the
= AUTOCORR(LAGS=) table is displayed by default.
option. This
Effective Sample Size
The “Effective Sample Sizes” table (ODS table name
ESS
) calculates the effective sample size of
each parameter. See the section “ Effective Sample Size ” on page 168 for more details. The table is
displayed by default.
Monte Carlo Standard Errors
The “Monte Carlo Standard Errors” table (ODS table name
MCSE
) calculates the standard errors of
the posterior mean estimate. See the section “ Standard Error of the Mean Estimate ” on page 169 for
more details. The table is displayed by default.
Geweke Diagnostics
The “Geweke Diagnostics” table (ODS table name
Geweke
) lists the result of the Geweke diagnostic
test. See the section “ Geweke Diagnostics ” on page 162 for more details. The table is displayed by
default.
Heidelberger-Welch Diagnostics
The “Heidelberger-Welch Diagnostics” table (ODS table name
Heidelberger
) lists the result of the
Heidelberger-Welch diagnostic test. The test is consisted of two parts: a stationary test and a half-
width test. See the section “ Heidelberger and Welch Diagnostics ” on page 164 for more details. The
table is not displayed by default and can be requested by specifying
Raftery-Lewis Diagnostics
The “Raftery-Lewis Diagnostics” table (ODS table name
Raftery
) lists the result of the Raftery-Lewis
diagnostic test. See the section “ Raftery and Lewis Diagnostics ” on page 165 for more details. The
table is not displayed by default and can be requested by specifying
Displayed Output
F
4199
Optimization Related ODS Tables
PROC MCMC can perform optimization on the joint posterior distribution. This is requested by the
option. The most commonly used optimization method is the quasi-Newton method:
PROPCOV=QUANEW(ITPRINT) . The ITPRINT option displays the ODS tables, listed as follows:
Input Options
The “Input Options” table (ODS table name procedure.
InputOptions
) lists optimization options used in the
Optimization Start
The “Optimization Start” table (ODS table name optimization.
ProblemDescription
) shows the initial state of the
Iteration History
The “Iteration History” table (ODS table name
IterHist
) shows iteration history of the optimization.
Optimization Results
The “Optimization Results” table (ODS table name
IterStop
) shows the results of the optimization, includes information about the number of function calls, and the optimized objective function, which is the joint log posterior density.
Convergence Status
The “Convergence Status” table (ODS table name
ConvergenceStatus
) shows whether the convergence criterion is satisfied.
Parameters Value After Optimization Table
The “Parameter Values After Optimization” table (ODS table name
OptiEstimates
) lists the parameter values that maximize the joint log posterior. These are the maximum a posteriori point estimates, and they are used to start the Markov chain.
Covariance Matrix After Optimization Table
The “Proposal Covariance” table (ODS table name
OptiCov
) lists covariance matrices for each block parameter by using quadrature approximation at the posterior mode. These covariance matrices are used in the proposal distribution.
4200
F
Chapter 52: The MCMC Procedure
PROC MCMC assigns a name to each table it creates. You can use these names to reference the table when using the Output Delivery System (ODS) to select tables and create output data sets. These
Table 52.35
ODS Tables Produced in PROC MCMC
ODS Table Name Description Statement or Option
autocorrelation statistics for each parameter basic statistics for each parameter, including sample size, mean, standard deviation, and percentiles optimization convergence status correlation matrix of the posterior samples covariance matrix of the posterior samples deviance information criterion effective sample size for each parameter
Monte Carlo standard error for each parameter
Geweke diagnostics for each parameter
Heidelberger-Welch diagnostics for each parameter optimization input table equal-tail and HPD intervals for each parameter optimization iteration history optimization results table number of observations parameter values after either optimization covariance used in proposal distribution after optimization summary of
BLOCKING, the PARMS,
PRIOR, sampling method, and initial value specification parameter values after the tuning phase default default
PROPCOV= method
(ITPRINT)
STATS=CORR
STATS=COV
DIC default default default
DIAGNOSTICS=HEIDEL
PROPCOV= method
(ITPRINT) default
PROPCOV= method
(ITPRINT)
PROPCOV= method
(ITPRINT) default
PROPCOV= method
(ITPRINT)
PROPCOV= method
(ITPRINT) default
INIT=PINIT
ODS Graphics
F
4201
Table 52.35
(continued)
ODS Table Name
Description posterior samples for each parameter optimization table
Raftery-Lewis diagnostics for each parameter history of burn-in and main phase sampling proposal covariance matrix (for continuous parameters) after the tuning phase proposal probability vector (for discrete parameters) after the tuning phase history of proposal distribution tuning
Statement or Option
(for ODS output data set only)
PROPCOV= method
(ITPRINT)
DIAGNOSTICS=RAFTERY default
INIT=PINIT
INIT=PINIT
CRETE=GEO default and DIS-
To request graphics with PROC MCMC, you must first enable ODS Graphics by specifying the ODS
GRAPHICS ON statement. See Chapter 21, “ Statistical Graphics Using ODS ,” for more information.
You can reference every graph produced through ODS Graphics with a name. The names of the graphs that PROC MCMC generates are listed in
Table 52.36
ODS Graphics Produced by PROC MCMC
ODS Graph Name
ADPanel
AutocorrPanel
AutocorrPlot
DensityPanel
DensityPlot
TAPanel
TADPanel
TDPanel
TracePanel
TracePlot
Plot Description autocorrelation function and density panel autocorrelation function panel autocorrelation function plot density panel density plot trace and autocorrelation function panel trace, density, and autocorrelation function panel trace and density panel trace panel trace plot
Statement & Option
PLOTS=(AUTOCORR DENSITY)
PLOTS=AUTOCORR
PLOTS(UNPACK)=AUTOCORR
PLOTS=DENSITY
PLOTS(UNPACK)=DENSITY
PLOTS=(TRACE AUTOCORR)
PLOTS=(TRACE AUTOCORR DENSITY)
PLOTS=(TRACE DENSITY)
PLOTS=TRACE
PLOTS(UNPACK)=TRACE
4202
F
Chapter 52: The MCMC Procedure
This example illustrates how you can obtain random samples from a known function. The target distributions are the normal distribution and a mixture of the normal distributions. You do not need any input data set to generate samples from a known density. You can set the likelihood function to a constant. The posterior distribution becomes identical to the prior distributions that you specify.
Sampling from a Normal Density
With a constant likelihood, there is no need to input a response variable since no data are relevant to a flat likelihood. However, PROC MCMC requires an input data set, so you can use an empty data set as the input data set. The following statements generate 10000 samples from a standard normal distribution:
title 'Simulating Samples from a Normal Density'; data x; run; ods graphics on; proc mcmc data=x outpost=simout seed=23 nmc=10000 maxtune=0 nbi=0 statistics=(summary interval) diagnostics=none; ods exclude nobs parameters samplinghistory; parm alpha 0; prior alpha ~ normal(0, sd=1); model general(0); run; ods graphics off;
The ODS GRAPHICS ON statement requests ODS Graphics. The PROC MCMC statement specifies the input and output data sets, a random number seed, and the size of the simulation sample. There
is no need for tuning ( MAXTUNE =0) because the default scale and the proposal variance are
The
option is used to display only the summary and interval statistics. The ODS
EXCLUDE statement excludes the display of the
and
tables. The
summary statistics ( Output 52.1.1
) are what you would expect from a standard normal distribution.
Example 52.1: Simulating Samples From a Known Density
F
4203
Output 52.1.1
MCMC Summary and Interval Statistics from a Normal Target Distribution
Parameter alpha
Simulating Samples from a Normal Density
The MCMC Procedure
N
10000
Posterior Summaries
Mean
Standard
Deviation
-0.0392
1.0194
25%
Percentiles
50%
-0.7198
-0.0403
75%
0.6351
Parameter alpha
Alpha
0.050
Posterior Intervals
Equal-Tail Interval
-2.0746
1.9594
HPD Interval
-2.2197
1.7869
The trace plot ( Output 52.1.2
) shows good mixing of the Markov chain, and there is no significant
autocorrelation in the lag plot.
Output 52.1.2
Diagnostics Plots for
˛
You can also overlay the estimated kernel density with the true density to get a visual comparison, as displayed in
4204
F
Chapter 52: The MCMC Procedure
To create
, you first use PROC KDE (see Chapter 45, “ The KDE Procedure ”) to obtain
a kernel density estimate of the posterior density on alpha
, and then you evaluate a grid of alpha values by using PROC KDE output data set sample on a normal density. The following statements evaluate kernel density and compute corresponding normal density.
proc kde data=simout; ods exclude inputs controls; univar alpha /out=sample; run; data den; set sample; alpha = value; true = pdf('normal', alpha, 0, 1); keep alpha density true; run;
Finally, you plot the two curves on top of each other by using PROC SGPLOT (see Chapter 21,
“ Statistical Graphics Using ODS ”); the resulting figure is in
estimate and the true density are very similar to one another. The following statements produce
proc sgplot data=den; yaxis label="Density"; series y=density x=alpha / legendlabel = "MCMC Kernel"; series y=true x=alpha / legendlabel = "True Density"; discretelegend; run;
Output 52.1.3
Estimated Density versus the True Density
Example 52.1: Simulating Samples From a Known Density
F
4205
Sampling from a Mixture of Normal Densities
Suppose that you are interested in generating samples from a three-component mixture of normal distributions, with the density specified as follows: p.˛/ D 0:3 . 3; D 2/ C 0:4 .2; D 1/ C 0:3 .10; D 4/
The following statements generate random samples from this mixture density:
title 'Simulating Samples from a Mixture of Normal Densities'; data x; run; ods graphics on; proc mcmc data=x outpost=simout seed=1234 nmc=30000; ods select TADpanel; parm alpha 0.3; lp = logpdf('normalmix', alpha, 3, 0.3, 0.4, 0.3, -3, 2, 10, 2, 1, 4); prior alpha ~ general(lp); model general(0); run; ods graphics off;
The ODS SELECT statement displays the diagnostic plots. All other tables, such as the
tables, are excluded. The PROC MCMC statement uses the input data set x
, saves output to the simout data set, sets a random number seed, and simulates 30,000 samples.
The lp assignment statement evaluates the log density of alpha at the mixture density, using the SAS function LOGPDF. The number 3 after alpha in the LOGPDF function indicates that the density is a three-component normal mixture. The following three numbers,
0:3
, in the mixture;
3
,
2
, and
10 are the means;
2
,
1
, and
4
0:4
, and
0:3
, are the weights are the standard deviations. The
statement assigns this log density function to alpha as its prior. Note that the
function interprets the density on the log scale, and not the original scale. Hence, you must use the LOGPDF function, not the PDF function.
displays the results. The kernel density clearly shows three modes.
4206
F
Chapter 52: The MCMC Procedure
Output 52.1.4
Plots of Posterior Samples from a Mixture Normal Distribution
Using the following set of statements similar to the previous example, you can overlay the estimated kernel density with the true density. The comparison is shown in
proc kde data=simout; ods exclude inputs controls; univar alpha /out=sample; run; data den; set sample; alpha = value; true = pdf('normalmix', alpha, 3, 0.3, 0.4, 0.3, -3, 2, 10, 2, 1, 4); keep alpha density true; run; proc sgplot data=den; yaxis label="Density"; series y=density x=alpha / legendlabel = "MCMC Kernel"; series y=true x=alpha / legendlabel = "True Density"; discretelegend; run;
Example 52.2: Box-Cox Transformation
F
4207
Output 52.1.5
Estimated Density versus the True Density
Box-Cox transformations ( Box and Cox 1964 ) are often used to find a power transformation of a
dependent variable to ensure the normality assumption in a linear regression model. This example illustrates how you can use PROC MCMC to estimate a Box-Cox transformation for a linear regression model. Two different priors on the transformation parameter prior and a discrete prior. You can estimate the probability of are considered: a continuous being 0 with a discrete prior but not with a continuous prior. The IF-ELSE statements are demonstrated in the example.
Using a Continuous Prior on
The following statements create a SAS data set with measurements of y
(the response variable) and x
(a single dependent variable):
4208
F
Chapter 52: The MCMC Procedure
title 'Box-Cox Transformation, with a Continuous Prior on Lambda'; data boxcox; input y x @@; datalines;
10.0
78.2
3.0
8.5
72.6
87.4
8.3
9.0
59.7
9.5
8.1
3.4
20.1
0.1
4.8
1.4
90.1
0.1
9.8
1.1
1.1
42.5
0.9
5.1
... more lines ...
2.6
;
1.8
58.6
7.9
81.2
8.1
37.2
6.9
The Box-Cox transformation of y takes on the form of: y./
D
( y 1 log
.y/ if if
¤
D
0
0:
I
The transformed response y./ is assumed to be normally distributed: y i
./ normal
.ˇ
0
C ˇ
1 x i
;
2
/
The likelihood with respect to the original response y i is as follows: f .y
i j ; ˇ;
2
; x i
/ / .y
i j ˇ
0
C ˇ
1 x i
;
2
/ J.; y i
/ where J.; y i
/ is the Jacobian:
J.; y/ D y i
1
1=y i if if
¤
D
0 I
0:
And on the log-scale, the Jacobian becomes: log .J.; y//
D
.
1/ log .y
i
/ log
.y
i
/ if if
¤
D
0 I
0:
There are four model parameters: on
ˇ and a gamma prior on
2
.
; ˇ D f ˇ
0
; ˇ
1 g ; and
2
. You can considering using a flat prior
To consider only power transformations ( a uniform prior from
2 to
2
) on
¤ 0
), you can use a continuous prior (for example,
. One issue with using a continuous prior is that you cannot estimate the probability of probability mass on the point
D
0
0 . To do so, you need to consider a discrete prior that places positive
. See “ Modeling D 0 ” on page 4212.
Example 52.2: Box-Cox Transformation
F
4209
The following statements fit a Box-Cox transformation model:
ods graphics on; proc mcmc data=boxcox nmc=50000 thin=10 propcov=quanew seed=12567 monitor=(lda); ods select PostSummaries PostIntervals TADpanel; parms beta0 0 beta1 0 lda 1 s2 1; beginnodata; prior beta: ~ general(0); prior s2 ~ gamma(shape=3, scale=2); prior lda ~ unif(-2,2); sd = sqrt(s2); endnodata; ys = (y**lda-1)/lda; mu = beta0+beta1*x; ll = (lda-1)*log(y)+lpdfnorm(ys, mu, sd); model general(ll); run;
The
option initializes the Markov chain at the posterior mode and uses the estimated inverse Hessian matrix as the initial proposal covariance matrix. The
option selects as the variable to report. The ODS SELECT statement displays the summary statistics table, the interval statistics table, and the diagnostic plots.
The
statement puts all four parameters,
ˇ
0
,
ˇ
1
, , and
2
, in a single block and assigns initial values to each of them. Three
statements specify previously stated prior distributions for these parameters. The assignment to better to place the transformation inside the computational time.
sd transforms a variance to a standard deviation. It is
and
statements to save
The assignment to the symbol regression mean and ll ys evaluates the Box-Cox transformation of is the log likelihood of the transformed variable y
, where mu is the ys
. Note that the log of the
Jacobian term is included in the calculation of ll
.
Summary statistics and interval statistics for lda are listed in
Output 52.2.1
Box-Cox Transformation
Parameter lda
Box-Cox Transformation, with a Continuous Prior on Lambda
The MCMC Procedure
Posterior Summaries
N
5000
Mean
Standard
Deviation
0.4702
0.0284
25%
0.4515
Percentiles
50%
0.4703
75%
0.4884
4210
F
Chapter 52: The MCMC Procedure
Output 52.2.1
continued
Parameter lda
Alpha
0.050
Posterior Intervals
Equal-Tail Interval
0.4162
0.5269
HPD Interval
0.4197
0.5298
The posterior mean of is
0:47
, with a
95
% equal-tail interval of interval. The prefered power transformation would be
0:5
Œ0:42; 0:53
(rounding and a similar HPD up to the square root transformation).
shows diagnostics plots for lda
. The chain appears to converge, and you can proceed to make inferences. The density plot shows that the posterior density is relatively symmetric around its mean estimate.
Output 52.2.2
Diagnostic Plots for
To verify the results, you can use PROC TRANSREG (see Chapter 91, “ The TRANSREG Procedure ”)
to find the estimate of .
proc transreg data=boxcox details pbo; ods output boxcox = bc; model boxcox(y / convenient lambda=-2 to 2 by 0.01) = identity(x); output out=trans; run;
Example 52.2: Box-Cox Transformation
F
4211
ods graphics off;
Output from PROC TRANSREG is shown in
produces a similar point estimate of
D
0:46 and
, and the 95% confidence interval is shown in
Output 52.2.3
Box-Cox Transformation Using PROC TRANSREG
Output 52.2.4
Estimates Reported by PROC TRANSREG
Box-Cox Transformation, with a Continuous Prior on Lambda
The TRANSREG Procedure
Model Statement Specification Details
Description Value Type DF Variable
Dep
Ind
1 BoxCox(y) Lambda Used
Lambda
Log Likelihood
0.5
0.46
-167.0
Conv. Lambda 0.5
Conv. Lambda LL -168.3
CI Limit
Alpha
Options
-169.0
0.05
Convenient Lambda Used
1 Identity(x) DF 1
4212
F
Chapter 52: The MCMC Procedure
The ODS data set bc contains the 95% confidence interval estimates produced by PROC TRANSREG.
This ODS table is rather large, and you want to see only the relevant portion. The following statements generate the part of the table that is important and display
proc print noobs label data=bc(drop=rmse); title2 'Confidence Interval'; where ci ne ' ' or abs(lambda - round(lambda, 0.5)) < 1e-6; label convenient = '00'x ci = '00'x; run;
The estimated 90% confidence interval is
Œ0:41; 0:51
, which is very close to the reported Bayesian credible intervals. The resemblance of the intervals is probably due to the noninformative prior that you used in this analysis.
Output 52.2.5
Estimated Confidence Interval on
Box-Cox Transformation, with a Continuous Prior on Lambda
Confidence Interval
Dependent Lambda R-Square
Log
Likelihood
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
BoxCox(y)
-2.00
-1.50
-1.00
-0.50
0.00
0.41
0.42
0.43
0.44
0.45
0.46
0.47
0.48
0.49
0.50
0.51
1.00
1.50
2.00
+
0.14
0.17
0.22
0.39
0.78
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.89
0.79
0.70
-1030.56
-810.50
-602.53
-415.56
-257.92
-168.40
-167.86
-167.46
-167.19
-167.05
-167.04
-167.16
-167.41
-167.79
-168.28
-168.89
-253.09
-345.35
-435.01
*
*
*
*
*
<
*
*
*
*
*
Modeling
D 0
With a continuous prior on , you can get only a continuous posterior distribution, and this makes the probability of Pr
.
D 0 j data
/ equal to
0 by definition. To consider
D 0 as a viable solution to the Box-Cox transformation, you need to use a discrete prior that places some probability mass on the point 0 and allows for a meaningful posterior estimate of Pr
.
D 0 j data
/
.
This example uses a simulation study where the data are generated from an exponential likelihood.
The simulation implies that the correct transformation should be the logarithm and should be 0.
Example 52.2: Box-Cox Transformation
F
4213
Consider the following exponential model: where y
D exp .x
C
/; normal
.0; 1/
. The transformed data can be fitted with a linear model: log
.y/ D x C
The following statements generate a SAS data set with a gridded x and corresponding y
:
title 'Box-Cox Transformation, Modeling Lambda = 0'; data boxcox; do x = 1 to 8 by 0.025; ly = x + normal(7); y = exp(ly); output; end; run;
The log-likelihood function, after taking the Jacobian into consideration, is as follows:
8 log p.y
i j ; x i
/ D
ˆ
<
ˆ
:
.
1/ log log .y
i
/
.y
1
2 i
/ log
1
2 log
2
C
.
.y
i
2
C
.
log
.y
i
/ x i
/
2
2
1/= x i
/
2
2
C C
2
C
C
1 if if
¤
D
0
0:
I where
C
1 and
C
2 are two constants.
You can use the function function
to place a discrete prior on . The function is similar to the
GENERAL , except that it indicates a discrete distribution. For example, you can specify a
discrete uniform prior from 2 to 2 using
prior lda ~ dgeneral(1, lower=-2, upper=2);
This places equal probability mass on five points,
2
,
1
,
0
,
1
, and well here because the grid is too coarse. To consider smaller values of
2
. This prior might not work
, you can sample a parameter that takes a wider range of integer values and transform it back to the as your model parameter and give it a discrete uniform prior from alpha/100 so can take values between
2 and
2 but on a finer grid.
space. For example, set
200 to
200
. Then define alpha as
The following statements fit a Box-Cox transformation by using a discrete prior on :
proc mcmc data=boxcox outpost=simout nmc=50000 thin=10 seed=12567 monitor=(lda); ods select PostSummaries PostIntervals; parms s2 1 alpha 10; beginnodata; prior s2 ~ gamma(shape=3, scale=2); if alpha=0 then lp = log(2); else lp = log(1); prior alpha ~ dgeneral(lp, lower=-200, upper=200); lda = alpha * 0.01; sd = sqrt(s2);
4214
F
Chapter 52: The MCMC Procedure
endnodata; if alpha=0 then ll = -ly+lpdfnorm(ly, x, sd); else do; ys = (y**lda - 1)/lda; ll = (lda-1)*ly+lpdfnorm(ys, x, sd); end; model general(ll); run;
There are two parameters, s2 and alpha
, in the model. They are placed in a single
statement so that they are sampled in the same block.
The parameter s2 takes a gamma distribution, and alpha takes a discrete prior. The IF-ELSE statements state that alpha takes twice as much prior density when it is 0 than otherwise. Note that on the original scale, Pr densities become log
.2/
.
alpha
D 0/ D 2
Pr
.
alpha and log
.1/
, respectively. The
¤ 0/ lda
. Translating that to the log scale, the assignment statement transforms alpha to the parameter of interest: lda takes values between 2 and 2 . You can model lda on a even smaller scale by dividing alpha by a larger constant. However, an increment of 0.01 in the Box-Cox transformation is usually sufficient. The sd assignment statement calculates the square root of the variance term.
The log-likelihood function uses another set of IF-ELSE statements, separating the case of
D 0 from the others. The formulas are stated previously. The output of the program is shown in
Output 52.2.6
Box-Cox Transformation
Parameter lda
Box-Cox Transformation, Modeling Lambda = 0
The MCMC Procedure
N
5000
Posterior Summaries
Mean
Standard
Deviation
-0.00002
0.00201
25%
0
Percentiles
50%
0
75%
0
Parameter lda
Alpha
0.050
Posterior Intervals
Equal-Tail Interval
0 0
HPD Interval
0 0
From the summary statistics table, you see that the point estimate for is 0 and both of the 95% equal-tail and HPD credible intervals are 0. This strongly suggests that this problem. In addition, you can also count the frequency of
D 0 is the best estimate for among posterior samples to get a more precise estimate on the posterior probability of being 0.
The following statements use PROC FREQ to produce
and
Example 52.2: Box-Cox Transformation
F
4215
ods graphics on; proc freq data=simout; ods select onewayfreqs freqplot; tables lda /nocum plot=freqplot(scale=percent); run; ods graphics off;
shows the frequency count table. An estimate of Pr .
D 0 j data / is 96 %. The conclusion is that the log transformation should be the appropriate transformation used here, which agrees with the simulation setup.
shows the histogram of .
Output 52.2.7
Frequency Counts of
Box-Cox Transformation, Modeling Lambda = 0
The FREQ Procedure lda Frequency Percent
---------------------------------
-0.0100
106 2.12
0
0.0100
4798
96
95.96
1.92
Output 52.2.8
Histogram of
4216
F
Chapter 52: The MCMC Procedure
This example discusses two examples of fitting generalized linear models (GLM) with PROC MCMC.
One uses a logistic regression model and one uses a Poisson regression model. The logistic examples use both a diffuse prior and a Jeffreys’ prior on the regression coefficients. You can also use the
BAYES statement in PROC GENMOD. See Chapter 37, “ The GENMOD Procedure .”
Logistic Regression Model with a Diffuse Prior
The following statements create a SAS data set with measurements of the number of deaths, y
, among n beetles that have been exposed to an environmental contaminant x
:
6
7
6
8
; title 'Logistic Regression Model with a Diffuse Prior'; data beetles; input n y x @@; datalines;
0
2
6
2
25.7
32.3
49.6
35.2
8
5
6
6
2
1
3
6
35.9
33.2
39.8
51.3
5
8
6
5
2
3
4
3
32.9
40.9
43.6
42.5
7
6
6
7
7
0
1
0
50.4
36.5
34.1
31.3
6
6
7
3
0
1
1
2
28.3
36.5
37.4
40.6
You can model the data points y i with a binomial distribution: y i j p i binomial
.n
i
; p i
/ where p i mation: is the success probability and links to the regression covariate x i through a logit transforlogit .p
i
/
D log
1 p i p i
D
˛
C
ˇx i
The priors on
˛ and
ˇ are both diffuse normal:
.˛/
.ˇ/
D
D
.0; var
D 10000/
.0; var D 10000/
These statements fit a logistic regression with PROC MCMC:
ods graphics on; proc mcmc data=beetles ntu=1000 nmc=20000 nthin=2 propcov=quanew diag=(mcse ess) outpost=beetleout seed=246810; ods select PostSummaries PostIntervals mcse ess TADpanel; parms (alpha beta) 0; prior alpha beta ~ normal(0, var = 10000); p = logistic(alpha + beta*x); model y ~ binomial(n,p); run;
Example 52.3: Generalized Linear Models
F
4217
The key statement in the program is the assignment to p that calculates the probability of death. The
SAS function LOGISTIC does the proper transformation. The
statement specifies that the response variable, y
, is binomially distributed with parameters n
(from the input data set) and p
The summary statistics table, interval statistics table, the Monte Carlos standard error table, and the
.
effective sample sizes table are shown in
Output 52.3.1
MCMC Results
Parameter alpha beta
Logistic Regression Model with a Diffuse Prior
The MCMC Procedure
Posterior Summaries
N
10000
10000
Mean
-11.7707
0.2920
Standard
Deviation
2.0997
0.0542
25%
-13.1243
0.2537
Percentiles
50%
-11.6683
0.2889
75%
-10.3003
0.3268
Parameter alpha beta
Alpha
0.050
0.050
Logistic Regression Model with a Diffuse Prior
The MCMC Procedure
Monte Carlo Standard Errors
Parameter alpha beta
MCSE
0.0422
0.00110
Standard
Deviation
2.0997
0.0542
MCSE/SD
0.0201
0.0203
Parameter alpha beta
Posterior Intervals
Equal-Tail Interval
-16.3332
0.1951
-7.9675
0.4087
Effective Sample Sizes
ESS
Autocorrelation
Time
2470.1
2435.4
4.0484
4.1060
HPD Interval
-15.8822
0.1901
-7.6673
0.4027
Efficiency
0.2470
0.2435
The summary statistics table shows that the sample mean of the output chain for the parameter alpha is
11:7707
. This is an estimate of the mean of the marginal posterior distribution for the intercept parameter
95 alpha
. The estimated posterior standard deviation for
% credible intervals for alpha alpha is 2.0997. The two are both negative, which indicates with very high probability that the intercept term is negative. On the other hand, you observe a positive effect on the regression coefficient beta
. Exposure to the environment contaminant increases the probability of death.
4218
F
Chapter 52: The MCMC Procedure
The Monte Carlo standard errors of each parameter are significantly small relative to the posterior standard deviations. A small MCSE/SD ratio indicates that the Markov chain has stabilized and the mean estimates do not vary much over time. Note that the precision in the parameter estimates increases with the square of the MCMC sample size, so if you want to double the precision, you must quadruple the MCMC sample size.
MCMC chains do not produce independent samples. Each sample point depends on the point before it. In this case, the correlation time estimate, read from the effective sample sizes table, is roughly 4.
This means that it takes four observations from the MCMC output to make inferences about alpha with the same precision that you would get from using an independent sample. The effective sample size of 2470 reflects this loss of efficiency. The coefficient beta has similar efficiency. You can often observe that some parameters have significantly better mixing (better efficiency) than others, even in a single Markov chain run.
Output 52.3.2
Plots for Parameters in the Logistic Regression Example
Output 52.3.2
continued
Example 52.3: Generalized Linear Models
F
4219
Trace plots and autocorrelation plots of the posterior samples are shown in
gence looks good in both parameters; there is good mixing in the trace plot and quick drop-off in the
ACF plot.
One advantage of Bayesian methods is the ability to directly answer scientific questions. In this example, you might want to find out the posterior probability that the environmental contaminant increases the probability of death—that is,
P r.ˇ > 0 j y /
. This can be estimated using the following steps:
proc format; value betafmt low-0 = 'beta <= 0' 0<-high = 'beta > 0'; run; proc freq data=beetleout; tables beta /nocum; format beta betafmt.; run;
4220
F
Chapter 52: The MCMC Procedure
Output 52.3.3
Frequency Counts
Logistic Regression Model with a Diffuse Prior
The FREQ Procedure beta Frequency Percent
---------------------------------beta > 0 10000 100.00
All of the simulated values for probability that
ˇ > 0
ˇ are greater than zero, so the sample estimate of the posterior is 100%. The evidence overwhelmingly supports the hypothesis that increased levels of the environmental contaminant increase the probability of death.
If you are interested in making inference based on any quantities that are transformations of the random variables, you can either do it directly in PROC MCMC or by using the DATA step after you run the simulation. Transformations sometimes can make parameter inference quite formidable using direct analytical methods, but with simulated chains, it is easy to compute chains for any set of parameters. Suppose that you are interested in the lethal dose and want to estimate the level of the covariate x that corresponds to a probability of death, p . Abbreviate this quantity as ldp
. In other words, you want to solve the logit transformation with a fixed value p
. The lethal dose is as follows: ldp
D log p
1 p
ˇ
˛
You can obtain an estimate of any ldp by using the posterior mean estimates for ˛ lp95
, which corresponds to p D 0:95 , is calculated as follows: and ˇ . For example, lp95
D log
0:95
1 0:95
0:29
C 11:77
D
50:79 where 11:77 and 0:29 are the posterior mean estimates of estimated lethal dose that leads to a
95
% death rate.
˛ and ˇ , respectively, and 50:79 is the
While it is easy to obtain the point estimates, it is harder to estimate other posterior quantities, such as the standard deviation directly. However, with PROC MCMC, you can trivially get estimates of any posterior quantities of lp95
. Consider the following program in PROC MCMC:
proc mcmc data=beetles ntu=1000 nmc=20000 nthin=2 propcov=quanew outpost=beetleout seed=246810 plot=density monitor=(pi30 ld05 ld50 ld95); ods select PostSummaries PostIntervals densitypanel; parms (alpha begincnst; beta) 0; c1 = log(0.05 / 0.95); c2 = -c1; endcnst; beginnodata; prior alpha beta ~ normal(0, var = 10000); pi30 = logistic(alpha + beta*30);
Example 52.3: Generalized Linear Models
F
4221
ld05 = (c1 - alpha) / beta; ld50 = - alpha / beta; ld95 = (c2 - alpha) / beta; endnodata; pi = logistic(alpha + beta*x); model y ~ binomial(n,pi); run; ods graphics off;
The program estimates four additional posterior quantities. The three lpd quantities, ld05
, ld50
, and ld95
, are the three levels of the covariate that kills 5%, 50%, and 95% of the population, respectively.
The predicted probability when the covariate x takes the value of 30 is pi30
. The
option selects the quantities of interest. The
option selects kernel density plots as the only ODS graphical output, excluding the trace plot and autocorrelation plot.
Programming statements between the
and
statements define two constants.
These statements are executed once at the beginning of the simulation. The programming statements between the
and
statements evaluate the quantities of interest. The symbols, pi30
, ld05
, ld50
, and ld95
, are functions of the parameters alpha and beta only. Hence, they should not be processed at the observation level and should be included in the
and
statements.
density plots of these posterior quantities.
lists the posterior summary and
shows the
Output 52.3.4
PROC MCMC Results
Parameter pi30 ld05 ld50 ld95
Logistic Regression Model with a Diffuse Prior
The MCMC Procedure
N
10000
10000
10000
10000
Posterior Summaries
Mean
Standard
Deviation
0.0524
29.9281
40.3745
50.8210
0.0253
1.8814
0.9377
2.5353
25%
Percentiles
50%
0.0340
28.8430
39.7271
49.0372
0.0477
30.1727
40.3165
50.5157
75%
0.0662
31.2563
40.9612
52.3100
Parameter pi30 ld05 ld50 ld95
Alpha
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
0.0161
25.6409
38.6706
46.7180
0.1133
32.9660
42.3718
56.7667
HPD Interval
0.0109
26.2193
38.6194
46.3221
0.1008
33.2774
42.2811
55.8774
The posterior mean estimate of lp95 is
50:82
, which is close to the estimate of
50:79 by using the posterior mean estimates of the parameters. With PROC MCMC, in addition to the mean estimate, you can get the standard deviation, quantiles, and interval estimates at any level of significance.
4222
F
Chapter 52: The MCMC Procedure
From the density plots, you can see, for example, that the sample distribution for the right, and almost all of your posterior belief concerning zero and 0.15.
30 is skewed to
30 is concentrated in the region between
Output 52.3.5
Density Plots of Quantities of Interest in the Logistic Regression Example
It is easy to use the DATA step to calculate these quantities of interest. The following DATA step uses the simulated values of
˛ and
ˇ to create simulated values from the posterior distributions of ld05, ld50, ld95, and
30
:
data transout; set beetleout; pi30 = logistic(alpha + beta*30); ld05 = (log(0.05 / 0.95) - alpha) / beta; ld50 = (log(0.50 / 0.50) - alpha) / beta; ld95 = (log(0.95 / 0.05) - alpha) / beta; run;
Subsequently, you can use SAS/INSIGHT, or the UNIVARIATE, CAPABILITY, or KDE procedures to analyze the posterior sample. If you want to regenerate the default ODS graphs from PROC
MCMC, see “ Regenerating Diagnostics Plots ” on page 4182.
Logistic Regression Model with Jeffreys’ Prior
A controlled experiment was run to study the effect of the rate and volume of air inspired on a transient reflex vasoconstriction in the skin of the fingers. Thirty-nine tests under various combinations of rate
Example 52.3: Generalized Linear Models
F
4223
and volume of air inspired were obtained ( Finney 1947 ). The result of each test is whether or not
vasoconstriction occurred.
Pregibon ( 1981 ) uses this set of data to illustrate the diagnostic measures
he proposes for detecting influential observations and to quantify their effects on various aspects of the maximum likelihood fit. The following statements create the data set vaso
:
3.7
0.8
0.9
0.6
3.2
0.4
1.6
title 'Logistic Regression Model with Jeffreys Prior'; data vaso; input vol rate resp @@; lvol = log(vol); lrate = log(rate); ind = _n_; cnst = 1; datalines;
0.825
3.2
0.75
3.0
1.6
2.0
1.78
1.9
1.1
0.95
1.83
0.95 1.9
;
1
1
0
0
1
0
1
1
0
0
3.5
0.7
0.9
1.4
1.6
1.1
1.09
3.5
0.45
2.33
0.4
2.2
0.75 1.9
1
1
0
1
0.85 1.415 1
0.95 1.36
0.6
1.5
0
0
0
1
0
1.25
0.6
0.8
0.75
1.7
1.35
1.8
2.7
1.2
1.3
2.5
0.75
0.57
3.75
1.06
1.35
1.5
0.75
2.0
1
0
0
1
0
0
1
1
1
1.625 1
0.75
1.1
0.55
2.3
1.8
1.5
0.8
1.5
1.7
1.64
1.8
1.36
0.95 1.9
2.35 0.03
3.33
1
0
2.75 0
1
1
0
0
0
1
The variable resp represents the outcome of a test. The variable lvol represents the log of the volume of air intake, and the variable lrate represents the log of the rate of air intake. You can model the data by using logistic regression. You can model the response with a binary likelihood: resp i binary .p
i
/ with p i
1
D
1 C exp
. .ˇ
0
C ˇ
1 lvol i
C ˇ
2 lrate i
//
Let
X be the design matrix in the regression. Jeffreys’ prior for this model is p.ˇ/ / j X
>
M
X j
1=2 where p i
.1
M is a 39 by 39 matrix with off-diagonal elements being 0 and diagonal elements being p i
/
. For details on Jeffreys’ prior, see “ Jeffreys’ Prior ” on page 144. You can use a number
of matrix functions, such as the determinant function, in PROC MCMC to construct Jeffreys’ prior.
The following statements illustrate how to fit a logistic regression with Jeffreys’ prior:
/* fitting a logistic regression with Jeffreys' prior */
%let n = 39; proc mcmc data=vaso nmc=10000 outpost=mcmcout seed=17; ods select PostSummaries PostIntervals; array beta[3] beta0 beta1 beta2; array m[&n, &n]; array x[&n, 3]; array xt[3, &n];
4224
F
Chapter 52: The MCMC Procedure
array xtm[3, &n]; array xmx[3, 3]; array p[&n]; parms beta0 1 beta1 1 beta2 1; begincnst; x[ind, 1] = 1; x[ind, 2] = lvol; x[ind, 3] = lrate; if (ind eq &n) then do; call transpose(x, xt); call zeromatrix(m); end; endcnst; beginnodata; call mult(x, beta, p); do i = 1 to &n; p[i] = 1 / (1 + exp(-p[i])); m[i,i] = p[i] * (1-p[i]); end; call mult (xt, m, xtm); call mult (xtm, x, xmx); call det (xmx, lp); lp = 0.5 * log(lp); prior beta: ~ general(lp); endnodata;
/* p = x * beta */
/* p[i] = 1/(1+exp(-x*beta)) */
/* xtm = xt * m
/* xmx = xtm * x
*/
*/
/* lp = det(xmx)
*/
/* lp = -0.5 * log(lp) */ model resp ~ bern(p[ind]); run;
The first
statement defines an array beta with three elements: beta0
, beta1
, and beta2
. The subsequent statements define arrays that are used in the construction of Jeffreys’ prior. These include m
(the
M matrix), x
(the design matrix), xt
(the transpose of x
), and some additional work spaces.
The explanatory variables lvol and lrate are saved in the array x in the
and
ENDCNST statements. See “ BEGINCNST/ENDCNST Statement ” on page 4133 for details. After all the
variables are read into x
, you transpose the x matrix and store it to xt
. The ZEROMATRIX function call assigns all elements in matrix m the value zero. To avoid redundant calculation, it is best to perform these calculations as the last observation of the data set is processed—that is, when ind is 39.
You calculate Jeffreys’ prior in the
vector p is the product of the design matrix x and
and parameter vector statements. The probability beta
. The diagonal elements in the matrix m are statement assigns lp p i
.1
p i
/ . The expression lp is the logarithm of Jeffreys’ prior. The as the prior for the
ˇ regression coefficients. The
statement assigns a binary likelihood to resp
, with probability p[ind]
. The p array is calculated earlier using the matrix function MULT. You use the ind variable to pick out the right probability value for each resp
.
Posterior summary statistics are displayed in
Example 52.3: Generalized Linear Models
F
4225
Output 52.3.6
PROC MCMC Results, Jeffreys’ prior
Parameter beta0 beta1 beta2
Logistic Regression Model with Jeffreys Prior
The MCMC Procedure
N
10000
10000
10000
Posterior Summaries
Mean
Standard
Deviation
-2.9587
5.2905
4.6889
1.3258
1.8193
1.8189
25%
Percentiles
50%
-3.8117
3.9861
3.3570
-2.7938
5.1155
4.4914
75%
-2.0007
6.4145
5.8547
Parameter beta0 beta1 beta2
Alpha
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-5.8247
2.3001
1.6788
-0.7435
9.3789
8.6643
HPD Interval
-5.5936
1.8590
1.3611
-0.6027
8.7222
8.2490
You can also use PROC GENMOD to fit the same model by using the following statements:
proc genmod data=vaso descending; ods select PostSummaries PostIntervals; model resp = lvol lrate / d=bin link=logit; bayes seed=17 coeffprior=jeffreys nmc=20000 thin=2; run;
The MODEL statement indicates that resp is the response variable and lvol and lrate are the covariates.
The options in the MODEL statement specify a binary likelihood and a logit link function. The
BAYES statement requests Bayesian capability. The SEED=, NMC=, and THIN= arguments work in the same way as in PROC MCMC. The COEFFPRIOR=JEFFREYS option requests Jeffreys’ prior in this analysis.
The PROC GENMOD statements produce
, with estimates very similar to those
reported in
. Note that you should not expect to see identical output from PROC
GENMOD and PROC MCMC, even with the simulation setup and identical random number seed.
The two procedures use different sampling algorithms. PROC GENMOD uses the adaptive rejection
metropolis algorithm (ARMS) ( Gilks and Wild 1992 ; Gilks 2003 ) while PROC MCMC uses a random
walk Metropolis algorithm. The asymptotic answers, which means that you let both procedures run an very long time, would be the same as they both generate samples from the same posterior distribution.
4226
F
Chapter 52: The MCMC Procedure
Output 52.3.7
PROC GENMOD Results
Parameter
Intercept lvol lrate
Logistic Regression Model with Jeffreys Prior
The GENMOD Procedure
Bayesian Analysis
Posterior Summaries
N
10000
10000
10000
Mean
Standard
Deviation
-2.8731
5.1639
4.5501
1.3088
1.8087
1.8071
25%
Percentiles
50%
-3.6754
3.8451
3.2250
-2.7248
4.9475
4.3564
75%
-1.9253
6.2613
5.6810
Parameter
Intercept lvol lrate
Alpha
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-5.8246
2.1844
1.5666
-0.7271
9.2297
8.6145
HPD Interval
-5.5774
2.0112
1.3155
-0.6060
8.9149
8.1922
Poisson Regression
You can use the Poisson distribution to model the distribution of cell counts in a multiway contingency table.
( 1989 ) have used this method to model insurance claims data. Suppose the
following hypothetical insurance claims data are classified by two factors: age group (with two levels) and car type (with three levels). The following statements create the data set:
title 'Poisson Regression'; data insure; input n c car $ age; ln = log(n); if car = 'large' then do car_dummy1=1; car_dummy2=0; end; else if car = 'medium' then do car_dummy1=0; car_dummy2=1; end; else do car_dummy1=0; car_dummy2=0; end; datalines;
500 42
1200 37 small 0 medium 0
100 1 large 0
Example 52.3: Generalized Linear Models
F
4227
400
500
300
;
101
73
14 small 1 medium 1 large 1
The variable n represents the number of insurance policy holders and the variable c represents the number of insurance claims. The variable car is the type of car involved (classified into three groups), and it is coded into two levels. The variable age is the age group of a policy holder (classified into two groups).
Assume that the number of claims related to the factors car and age c has a Poisson probability distribution and that its mean, for observation i by i
, is log .
i
/
D
D log .
n i
/
C x
0
ˇ log .
n i
/
C
ˇ
0
C car i
.1/ˇ
1
C car i
.2/ˇ
2
C car i
.3/ˇ
3
C age i
.1/ˇ
4
C age i
.2/ˇ
5
The indicator variables car i
.j / is associated with the in the following way: j th level of the variable car for observation i car i
.j / D
1
0 if car if car
D
¤ j j
A similar coding applies to age
. The
ˇ
’s are parameters. The logarithm of the variable n is used as an offset—that is, a regression variable with a constant coefficient of 1 for each observation. Having the offset constant in the model is equivalent to fitting an expanded data set with 3000 observations, each with response variable y observed on an individual level. The log link is used to relate the mean and the factors car and age
.
The following statements run PROC MCMC:
proc mcmc data=insure outpost=insureout nmc=5000 propcov=quanew maxtune=0 seed=7; ods select PostSummaries PostIntervals; parms alpha 0 beta_car1 0 beta_car2 0 beta_age 0; prior alpha beta: ~ normal(0, prec = 1e-6); mu = ln + alpha + beta_car1 * car_dummy1
+ beta_car2 * car_dummy2 + beta_age * age; model c ~ poisson(exp(mu)); run;
The analysis uses a relatively flat prior on all the regression coefficients, with mean at precision at
10
6
. The option
0 and skips the tuning phase because the optimization routine
( PROPCOV=QUANEW ) provides good initial values and proposal covariance matrix.
There are four parameters in the model: alpha is the intercept; beta_car1 and beta_car2 are coefficients for the class variable car
, which has three levels; and beta_age is the coefficient for age
. The symbol mu connects the regression model and the Poisson mean by using the log link. The
statement specifies a Poisson likelihood for the response variable c
.
4228
F
Chapter 52: The MCMC Procedure
Posterior summary and interval statistics are shown in
Output 52.3.8
MCMC Results
Parameter alpha beta_car1 beta_car2 beta_age
N
5000
5000
5000
5000
Poisson Regression
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-2.6403
-1.8335
-0.6931
1.3151
0.1344
0.2917
0.1255
0.1386
25%
Percentiles
50%
-2.7261
-2.0243
-0.7775
1.2153
-2.6387
-1.8179
-0.6867
1.3146
75%
-2.5531
-1.6302
-0.6118
1.4094
Parameter alpha beta_car1 beta_car2 beta_age
Alpha
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-2.9201
-2.4579
-0.9462
1.0442
-2.3837
-1.3036
-0.4498
1.5898
HPD Interval
-2.9133
-2.4692
-0.9485
1.0387
-2.3831
-1.3336
-0.4589
1.5812
To fit the same model by using PROC GENMOD, you can do the following. Note that the default normal prior on the coefficients ˇ is N.0; prec
D
1e 6/ , the same as used in the PROC MCMC.
The following statements run PROC GENMOD and create
proc genmod data=insure; ods select PostSummaries PostIntervals; class car age(descending); model c = car age / dist=poisson link=log offset=ln; bayes seed=17 nmc=5000 coeffprior=normal; run;
To compare, posterior summary and interval statistics from PROC GENMOD are reported in
, and they are very similar to PROC MCMC results in Output 52.3.8
Example 52.4: Nonlinear Poisson Regression Models
F
4229
Output 52.3.9
PROC GENMOD Results
Parameter
Intercept carlarge carmedium age1
N
5000
5000
5000
5000
Poisson Regression
The GENMOD Procedure
Bayesian Analysis
Posterior Summaries
Mean
Standard
Deviation
-2.6353
-1.7996
-0.6977
1.3148
0.1299
0.2752
0.1269
0.1348
25%
Percentiles
50%
-2.7243
-1.9824
-0.7845
1.2237
-2.6312
-1.7865
-0.6970
1.3138
75%
-2.5455
-1.6139
-0.6141
1.4067
Parameter
Intercept carlarge carmedium age1
Alpha
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-2.8952
-2.3538
-0.9494
1.0521
-2.3867
-1.2789
-0.4487
1.5794
HPD Interval
-2.8755
-2.3424
-0.9317
1.0624
-2.3730
-1.2691
-0.4337
1.5863
Note that the descending option in the CLASS statement reverses the sorting order of the class variable age so that the results agree with PROC MCMC. If this option is not used, the estimate for age has a reversed sign as compared to
This example illustrates how to fit a nonlinear Poisson regression with PROC MCMC. In addition, it shows how you can improve the mixing of the Markov chain by selecting a different proposal distribution or by sampling on the transformed scale of a parameter. This example shows how to analyze count data for calls to a technical support help line in the weeks immediately following a product release. This information could be used to decide upon the allocation of technical support resources for new products. You can model the number of daily calls as a Poisson random variable, with the average number of calls modeled as a nonlinear function of the number of weeks that have elapsed since the product’s release. The data are input into a SAS data set as follows:
1
4
7
; title 'Nonlinear Poisson Regression'; data calls; input weeks calls @@; datalines;
0
5
24
1
4
7
2
8
16
2
5
8
2
5
23
2
5
8
1
9
27
3
6
1
17
3
6
3
9
4230
F
Chapter 52: The MCMC Procedure
During the first several weeks after a new product is released, the number of questions that technical support receives concerning the product increases in a sigmoidal fashion. The expression for the mean value in the classic Poisson regression involves the log link. There is some theoretical justification for this link, but with MCMC methodologies, you are not constrained to exploring only models that are computationally convenient. The number of calls to technical support tapers off after the initial release, so in this example you can use a logistic-type function to model the mean number of calls received weekly for the time period immediately following the initial release. The mean function
.t / is modeled as follows: i
D
1 C exp
Œ .˛ C ˇt i
/
The likelihood for every observation calls i is calls i
Poisson
.
i
/
Past experience with technical support data for similar products suggests using a gamma distribution with shape and scale parameters 3.5 and 12 as the prior distribution for , a normal distribution with mean 5 and variance 0.25 as the prior for
0.5 as the prior for
ˇ
.
˛ , and a normal distribution with mean 0.75 and variance
The following PROC MCMC statements fit this model:
ods graphics on; proc mcmc data=calls outpost=callout seed=53197 ntu=1000 nmc=20000 propcov=quanew; ods select TADpanel; parms alpha -4 beta 1 gamma 2; prior alpha ~ normal(-5, sd=0.25); prior beta ~ normal(0.75, sd=0.5); prior gamma ~ gamma(3.5, scale=12); lambda = gamma*logistic(alpha+beta*weeks); model calls ~ poisson(lambda); run;
The one
statement defines a block of all parameters and sets their initial values individually.
The
statements specify the informative prior distributions for the three parameters. The assignment statement defines , the mean number of calls. Instead of using the SAS function
LOGISTIC, you can use the following statement to calculate and get the same result:
lambda = gamma / (1 + exp(-(alpha+beta*weeks)));
Mixing is not particularly good with this run of PROC MCMC. The ODS SELECT statement displays only the diagnostic graphs while excluding all other output. The graphical output is shown in
Example 52.4: Nonlinear Poisson Regression Models
F
4231
Output 52.4.1
Plots for Parameters
4232
F
Chapter 52: The MCMC Procedure
Output 52.4.1
continued
By examining the trace plot of the gamma parameter, you see that the Markov chain sometimes gets stuck in the far right tail and does not travel back to the high density area quickly. This effect can be seen around the simulations number 8000 and 18000. One possible explanation for this is that the random walk Metropolis is taking too small of steps in its proposal; therefore it takes more iterations for the Markov chain to explore the parameter space effectively. The step size in the random walk is controlled by the normal proposal distribution (with a multiplicative scale). A (good) proposal distribution is roughly an approximation to the joint posterior distribution at the mode. The curvature of the normal proposal distribution (the variance) does not take into account the thickness of the tail areas. As a result, a random walk Metropolis with normal proposal can have a hard time exploring t distributions that have thick tails. This appears to be the case with the posterior distribution of the parameter gamma
. You can improve the mixing by using a thicker-tailed proposal distribution, the
-distribution. The option
controls the proposal distribution.
changes the proposal from a normal distribution to a t -distribution with three degrees of freedom.
The following statements run PROC MCMC and produce
proc mcmc data=calls outpost=callout seed=53197 ntu=1000 nmc=20000 propcov=quanew stats=none propdist=t(3); ods select TADpanel; parms alpha -4 beta 1 gamma 2; prior alpha ~ normal(-5, sd=0.25); prior beta ~ normal(0.75, sd=0.5); prior gamma ~ gamma(3.5, scale=12); lambda = gamma*logistic(alpha+beta*weeks); model calls ~ poisson(lambda); run;
Example 52.4: Nonlinear Poisson Regression Models
F
4233
displays the graphical output.
Output 52.4.2
Plots for Parameters, Using a t(3) Proposal Distribution
4234
F
Chapter 52: The MCMC Procedure
Output 52.4.2
continued
The trace plots are more dense and the ACF plots have faster drop-offs, and you see improved mixing by using a thicker-tailed proposal distribution. If you want to further improve the Markov chain, you can choose to sample the log transformation of the parameter gamma
: lg egamma
.3:5; scale
D 12/ is equivalent to gamma D exp
.
lg
/ gamma
.3:5; scale
D 12/
The parameter gamma has a positive support. Often in this case, it has right-skewed posterior. By taking the log transformation, you can sample on a parameter space that does not have a lower boundary and is more symmetric. This can lead to better mixing.
The following statements produce
and
proc mcmc data=calls outpost=callout seed=53197 ntu=1000 nmc=20000 propcov=quanew propdist=t(3) monitor=(alpha beta lgamma gamma); ods select PostSummaries PostIntervals TADpanel; parms alpha -4 beta 1 lgamma 2; prior alpha ~ normal(-5, sd=0.25); prior beta ~ normal(0.75, sd=0.5); prior lgamma ~ egamma(3.5, scale=12); gamma = exp(lgamma); lambda = gamma*logistic(alpha+beta*weeks); model calls ~ poisson(lambda); run; ods graphics off;
Example 52.4: Nonlinear Poisson Regression Models
F
4235
In the
statement, instead of gamma
, you have lgamma
. Its prior distribution is egamma, as opposed to the gamma distribution. Note that the following two priors are equivalent to each other:
prior lgamma ~ egamma(3.5, scale=12); prior gamma ~ gamma(3.5, scale=12);
The gamma assignment statement transforms lgamma to gamma
. The lambda assignment statement calculates the mean for the Poisson by using the gamma parameter. The
statement specifies a Poisson likelihood for the calls response.
The trace plots and ACF plots in
show the best mixing seen so far in this example.
Output 52.4.3
Plots for Parameters, Sampling on the Log Scale of
Gamma
4236
F
Chapter 52: The MCMC Procedure
Output 52.4.3
continued
Output 52.4.3
continued
Example 52.4: Nonlinear Poisson Regression Models
F
4237
shows the posterior summary statistics of the nonlinear Poisson regression. Note that the lgamma parameter has a more symmetric density than the skewed gamma parameter. The
Metropolis algorithm always works better if the target distribution is approximately normal.
Output 52.4.4
MCMC Results, Sampling on the Log Scale of Gamma
Parameter alpha beta lgamma gamma
N
20000
20000
20000
20000
Nonlinear Poisson Regression
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-4.8907
0.6957
3.7391
44.8136
0.2160
0.1089
0.3487
17.0430
25%
Percentiles
50%
-5.0435
0.6163
3.4728
32.2263
-4.8872
0.6881
3.7023
40.5415
75%
-4.7461
0.7698
3.9696
52.9647
4238
F
Chapter 52: The MCMC Procedure
Output 52.4.4
continued
Parameter alpha beta lgamma gamma
Alpha
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-5.3138
0.5066
3.1580
23.5225
-4.4667
0.9253
4.4705
87.3972
HPD Interval
-5.3276
0.4868
3.1222
20.9005
-4.4953
0.8996
4.4127
79.4712
This example illustrates that PROC MCMC can fit Bayesian nonlinear models just as easily as
Bayesian linear models. More importantly, transformations can sometimes improve the efficiency of
the Markov chain, and that is something to always keep in mind. Also see “ Example 52.12: Using a
Transformation to Improve Mixing ” on page 4307 for another example of how transformations can
improve mixing of the Markov chains.
This example illustrates how you can use PROC MCMC to fit random effects models. In the example
” on page 4116 in “ Getting Started: MCMC Procedure ” on page 4103, you
already saw PROC MCMC fit a linear random effects model. There are two more examples in this section. One is a logistic random effects model, and the second one is a nonlinear Poisson regression random effects model. In addition, this section illustrates how to construct prior distributions that depend on input data set variables. Such prior distributions appear frequently in random effects model, especially in cases of hierarchical centering. Although you can use PROC MCMC to analyze random effects models, you might want to first consider some other SAS procedures. For example,
generalized linear mixed effects models. In addition, a sampling-based Bayesian analysis is available
in the MIXED procedure through the PRIOR statement (see “ PRIOR Statement ” on page 4569).
Logistic Regression Random-Effects Model
This example shows how to fit a logistic random-effects model in PROC MCMC. The data are taken from
seeds data set is a
2 2 factorial layout, with two types of seeds, O.
aegyptiaca 75 and O. aegyptiaca 73 , and two root extracts, which is the number of germinated seeds, and bean and cucumber . You observe r
, n
, which is the total number of seeds. The independent variables are seed and extract
.
Example 52.5: Random-Effects Models
F
4239
The following statements create the data set:
10
17
32
10 title 'Logistic Regression Random-Effects Model'; data seeds; input r n seed extract @@; ind = _N_; datalines;
39 0 0
39
51
0
0
0
1
30
12
7
1
1
1
0
1
1
23
5
46
8
22
62
6
79
28
41
0
0
0
1
1
0
1
1
0
1
23
53
10
23
15
81
74
13
45
30
0
0
0
1
1
0
1
1
0
1
26
55
8
0
32
;
3
3
51
72
16
4
51
0
0
1
1
1
0
1
0
0
1
You can model each observation r i as follows: as having its own probability of success p i
, and the likelihood is r i binomial
.
n i
; p i
/
You can use the logit link function to link the covariates of each observation, seed and extract
, to the probability of success: p i i
D
D
ˇ
0
C ˇ
1 logistic
.
i seed i
C ˇ
2
C i
/ extract i
C ˇ
3 seed i extract i where i i is assumed to be as i.i.d. random effect with a normal prior: normal
.0; var
D
2
/
The four
ˇ regression coefficients and the standard deviation parameters; they are given noninformative priors as follows:
2 in the random effects are model
.ˇ
0
; ˇ
1
; ˇ
2
; ˇ
3
/
.
2
/
/
/
1
1=
2
Another way of expressing the same model is as follows: p i
D logistic .ı i
/ where ı i normal
.ˇ
0
C ˇ
1 seed i
C ˇ
2 extract i
C ˇ
3 seed i extract i
;
2
/
The two models are equivalent. In the first model, the random effects distribution, and in the second model, ı i i centers at 0 in the normal centers at the regression mean. This hierarchical centering can sometimes improve mixing.
From a programming point of view, the second parameterization of the model is more difficult because the prior distribution on ı i involves the data set variables seed and extract
. Each prior distribution depends on a different set of observations in the input data set. Intuitively, you might think that the following statements would specify such a prior:
4240
F
Chapter 52: The MCMC Procedure
mu = beta0 + beta1*seed + beta2*extract + beta3*seed*extract; prior delta ~ normal(mu, var = v);
However, this will not work. This is because the procedure is not able to match the observational level calculation ( mu
) with elements of a parameter array (there are 21 random effects in delta
). Thus, the procedure cannot calculate the log of the prior density correctly. The solution is to cumulatively calculate the joint prior distribution for all ı i
; i D 1 21
, and assign the prior distribution to all ı by using the
function.
The following statements generate
proc mcmc data=seeds outpost=postout seed=332786 nmc=100000 thin=10 ntu=3000 monitor=(beta0-beta3 v); ods select PostSummaries ess; array delta[21]; parms delta: 0; parms beta0 0 beta1 0 beta2 0 beta3 0 ; parms v 1; beginnodata; sigma = sqrt(v); endnodata; w = beta0 + beta1*seed + beta2*extract + beta3*seed*extract; if ind eq 1 then lp = lpdfnorm(delta[ind], w, sigma); else lp = lp + lpdfnorm(delta[ind], w, sigma); prior v ~ general(-log(v)); prior beta: ~ general(0); prior delta: ~ general(lp); pi = logistic(delta[ind]); model r ~ binomial(n = n, p = pi); run;
PROC MCMC statement specifies the input and output data sets, sets a seed for the random number generator, requests a very large simulation number, thins the Markov chain by 10, and specifies a tuning sample size of 3000. The
option selects the parameters of interest. The ODS
SELECT statement displays the summary statistics and effective sample size tables.
The three
statement allocates an array of size 21 for the random effects parameter statements that place ı , ˇ and
2 ı
. There are into three sampling blocks. Calculation of sigma does not involve any observations; hence, it is enclosed in the
and
statements.
Example 52.5: Random-Effects Models
F
4241
The next few lines of statements construct a joint prior distribution for all the ı parameters. The symbol w is the regression mean, whose value changes for every observation. The IF-ELSE statements add the log of the normal density to the symbol lp as PROC MCMC steps through the data set. When ind is 1, lp is the log of the normal density for delta[1] evaluated at the first regression mean w
. As ind gradually increases to
21
, lp becomes
X log ..ı i j
ˇX i
; // i which is the joint prior distribution for all ı
.
The
ˇ
. All of the statements assign three priors to these parameters, with noninformative priors on delta parameters share a joint prior, which is defined by lp
2 and
. Recall that PROC MCMC adds the log of the prior density to the log of the posterior density at the last observation at every simulation, so the expression lp will have the correct value.
C
AUTION
: You must define the expression lp before the
statement for the delta parameters.
Switching the order of the
statement and the programming statements that define lp leads to an incorrect prior distribution for delta
. The following statements are wrong because the expression lp has not completed its calculation when observation of the input data set.
lp is added to the log of the posterior density at the last
prior delta: ~ general(lp); w = beta0 + beta1*seed + beta2*extract + beta3*seed*extract; if ind eq 1 then lp = lpdfnorm(delta[ind], w, sigma); else lp = lp + lpdfnorm(delta[ind], w, sigma);
The prior you specify in this case is: n 1
X log
..ı i j ˇX i
; // i D 1
The correct log density is the following: n
X log
..ı i j ˇX i
; // i D 1
The symbol pi is the logit transformation. The
specifies the response variable r as a binomial distribution with parameters n and pi
.
The mixing is poor in this example. You can see from the effective sample size table ( Output 52.5.1
that the efficiency for all parameters is relatively low, even after a substantial amount of thinning.
One possible solution is to break the random effects block of parameters ( b
) into multiple blocks with a smaller number of parameters.
4242
F
Chapter 52: The MCMC Procedure
Output 52.5.1
Logistic Regression Random-Effects Model
Parameter beta0 beta1 beta2 beta3 v
N
10000
10000
10000
10000
10000
Logistic Regression Random-Effects Model
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-0.5503
0.0626
1.3546
-0.8257
0.1145
0.2025
0.3292
0.2876
0.4498
0.1019
25%
Percentiles
50%
-0.6784
-0.1512
1.1732
-1.1044
0.0472
-0.5522
0.0653
1.3391
-0.8255
0.0875
75%
-0.4193
0.2760
1.5349
-0.5344
0.1503
Parameter beta0 beta1 beta2 beta3 v
Logistic Regression Random-Effects Model
The MCMC Procedure
Effective Sample Sizes
ESS
885.3
603.2
854.9
591.6
273.1
Autocorrelation
Time
11.2952
16.5771
11.6970
16.9021
36.6182
Efficiency
0.0885
0.0603
0.0855
0.0592
0.0273
To fit the same model in PROC GLIMMIX, you can use the following statements, which produce
proc glimmix data=seeds method=quad; ods select covparms parameterestimates; ods output covparms=cp parameterestimates=ps; class ind; model r/n = seed extract seed*extract/ dist=binomial link=logit solution; random intercept / subject=ind; run;
Output 52.5.2
Estimates by PROC GLMMIX
Logistic Regression Random-Effects Model
The GLIMMIX Procedure
Covariance Parameter Estimates
Cov Parm
Intercept
Subject ind
Estimate
0.05577
Standard
Error
0.05196
Example 52.5: Random-Effects Models
F
4243
Output 52.5.2
continued
Effect
Intercept seed extract seed*extract
Solutions for Fixed Effects
Estimate
Standard
Error DF t Value
-0.5484
0.09701
1.3370
-0.8104
0.1666
0.2780
0.2369
0.3851
17
0
0
0
-3.29
0.35
5.64
-2.10
Pr > |t|
0.0043
.
.
.
It is hard to compare point estimates from these two procedures. However, you can visually compare the results by plotting a kernel density plot (by using the posterior sample from PROC MCMC output) on top of a normal approximation plot (by using the mean and standard error estimates from
PROC GLIMMIX, for each parameter). This kernel comparison plot is shown in
However, it takes some work to produce the kernel comparison plot. First, you must use PROC
KDE to estimate the kernel density for each parameter from MCMC. Next, you want to get the point estimates from the PROC GLIMMIX output. Then, you generate a SAS data set that contains both the kernel density estimates and the gridded estimates based on normal approximations. Finally,
you use PROC TEMPLATE (see Chapter 21, “ Statistical Graphics Using ODS ”) to define an
appropriate graphical template and produce the comparison plot by using PROC SGRENDER (see the SGRENDER Procedure in the SAS/GRAPH: Statistical Graphics Procedures Guide ).
The following statements use PROC KDE on the posterior sample data set postout and estimate a kernel density for each parameter, saving the estimates to a SAS data set m1 :
proc kde data=postout; univar beta0 beta1 beta2 beta3 v / out=m1 (drop=count); run;
The following SAS statements take the estimates of all the parameters from the PROC GLIMMIX output, data sets ps and cp
, and assign them to macro variables:
data gmxest(keep = parm mean sd); set ps cp; mean = estimate; sd = stderr; i = _n_-1; if(_n_ ne 5) then parm = "beta" || put(i, z1.); else parm = "var"; run; data msd (keep=mean sd); set gmxest; do j = 1 to 401; output; end; run;
4244
F
Chapter 52: The MCMC Procedure
data _null_; set ps; call symputx(compress(effect,'*'), estimate); call symputx(compress('s' || effect,'*'), stderr); run; data _null_; set cp; call symputx("var", estimate); call symputx("var_sd", stderr); run;
%put &intercept &seed &extract &seedextract &var;
%put &sintercept &sseed &sextract &sseedextract &var_sd;
Specifically, the mean estimate of
ˇ
0 to sintercept
. The macro variables is assigned to seed
, extract
, intercept seedextract
, and the standard error of are the mean estimates for
ˇ
0
ˇ
1
, is assigned
ˇ
2 and ˇ
3
, respectively.
To create a SAS data set that contains both the kernel density estimates and the corresponding normal approximation, you can use the %REN and %RESHAPE macros. The %REN macro renames the variables of a SAS data set by appending the suffix name to each variable name, to avoid redundant variable names. The %RESHAPE macro takes an output data set from a PROC KDE run, and transposes it to the right format so that PROC SGRENDER can generate the right graph. The following statements define the %REN and %RESHAPE macros:
/* define macros */
%macro ren(in=, out=, suffix=);
%local s; proc contents data=&in noprint out=__temp__(keep=name); run; data _null_; length s $ 32000; retain s; set __temp__ end=eof; s = trim(s)||' '||trim(name)||'='||compress(name||"&suffix"); if eof then call symput('s', trim(s)); run; proc datasets nolist; delete __temp__; run; quit; data &out; set &in(rename=(&s)); run;
%mend;
%macro reshape(input, output, suffix1=, suffix2=); proc sort data=&input;
Example 52.5: Random-Effects Models
F
4245
by var; run; data tmp&input; set &input; by var;
_n + 1; if first.var then _n = 0; run; proc sort; by _n var; run; proc transpose data=tmp&input out=_by_value_(drop=_n _name_ _label_); var value; by _n; id var; run;
%ren(in=_by_value_, out=_by_value_, suffix=&suffix1) proc transpose data=tmp&input out=_by_den_(drop=_n _name_ _label_); var density; by _n; id var; run;
%ren(in=_by_den_, out=_by_den_, suffix=&suffix2) data &output; merge _by_value_ _by_den_; run; proc datasets library=work; ods exclude all; delete tmp&input _by_value_ _by_den_; run; ods exclude none;
%mend;
When you apply the %RESHAPE macro to the data set has grid values of the
ˇ m1
, you create a SAS data set mcmc that parameters and their corresponding kernel density estimates. Next, you evaluate these parameter grid values in a normal density with the macro variables taken from the
PROC GLIMMIX output:
/* create data set mcmc */
%reshape(m1, mcmc, suffix1=, suffix2=_kde); data all; set mcmc; beta0_gmx = pdf('normal', beta0, &intercept, &sintercept); beta1_gmx = pdf('normal', beta1, &seed, &sseed); beta2_gmx = pdf('normal', beta2, &extract, &sextract); beta3_gmx = pdf('normal', beta3, &seedextract, &sseedextract);
4246
F
Chapter 52: The MCMC Procedure
v_gmx = pdf('normal', v, &var, &var_sd); run;
In the data set all , you have grid values on
ˇ and
2
, their kernel density estimates from PROC
MCMC, and the normal density evaluated by using estimates from PROC GLIMMIX. To create an overlaid plot, you first use PROC TEMPLATE to create a 2 3 template as demonstrated by the following statements:
proc template; define statgraph twobythree;
%macro plot; begingraph; layout lattice / rows=2 columns=3;
%do i = 0 %to 3; layout overlay /yaxisopts=(label=" "); seriesplot y=beta&i._kde x=beta&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=beta&i._gmx x=beta&i
/ connectorder=xaxis lineattrs=(color=red) legendlabel="GLIMMIX Approximation" name="GLIMMIX"; endlayout;
%end; layout overlay /yaxisopts=(label=" ") xaxisopts=(linearopts=(viewmin=0 viewmax=0.6)); seriesplot y=v_kde x=v
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=v_gmx x=v
/ connectorder=xaxis lineattrs=(color=red) legendlabel="GLIMMIX Approximation" name="GLIMMIX"; endlayout;
Sidebar / align = bottom; discretelegend "MCMC" "GLIMMIX"; endsidebar; endlayout; end; run; endgraph;
%mend; %plot;
The kernel density comparison plot is produced by calling PROC SGRENDER (see the SGRENDER
Procedure in the SAS/GRAPH: Statistical Graphics Procedures Guide ):
proc sgrender data=all template=twobythree; run;
Example 52.5: Random-Effects Models
F
4247
Output 52.5.3
Comparing Estimates from PROC MCMC and PROC GLIMMIX.
The kernel densities are very similar to each other. Kernel densities from PROC MCMC are not as smooth, possibly due to bad mixing of the Markov chains.
Nonlinear Poisson Regression Random-Effects Model
This example uses the pump failure data of
Gaver and O’Muircheartaigh ( 1987 ). The number of
failures and the time of operation are recorded for 10 pumps. Each of the pumps is classified into one of two groups corresponding to either continuous or intermittent operation. The following statements generate the data set:
title 'Nonlinear Poisson Regression Random Effects Model'; data pump; input y t group @@; pump = _n_; logtstd = log(t) - 2.4564900; datalines;
5 94.320 1
14 125.760 1
1.048 2
10.480 2
1
3
1
15.720 2
5.240 2
1.048 2
5
19
4
;
1
22
62.880 1
31.440 1
2.096 2
4248
F
Chapter 52: The MCMC Procedure
Each row denotes data for a single pump, and the variable logtstd contains the centered operation times. Letting y ij denote the number of failures for the considers the following hierarchical model for these data: j th pump in the i th group,
y ij j ij log e ij j ij
2
D
Poisson
.
ij
/
˛ i
C ˇ i
.
log t ij normal
.0;
2
/ log t / C e ij
The model specifies different intercepts and slopes for each group, and the random effect is a mechanism for accounting for over-dispersion. You can use noninformative priors on the parameters
˛ i
,
ˇ i
, and
2
.
.˛
1
; ˛
2
; ˇ
1
; ˇ
2
/
.
2
/
/
/
1
1=
2
The following statements fit this nonlinear hierarchical model and produce
proc mcmc data=pump outpost=postout seed=248601 nmc=100000 ntu=2000 thin=10 monitor=(logsig beta1 beta2 alpha1 alpha2 s2 adif bdif); ods select PostSummaries; array alpha[2]; array beta[2]; array llambda[10]; parms (alpha: beta:) 1; parms llambda: 1; parms s2 1; beginnodata; sd = sqrt(s2); logsig = log(s2)/2; adif = alpha1 - alpha2; bdif = beta1 - beta2; endnodata; w = alpha[group] + beta[group] * logtstd; if pump eq 1 then lp = lpdfnorm(llambda[pump], w, sd); else lp = lp + lpdfnorm(llambda[pump], w, sd); prior alpha: beta: ~ general(0); prior s2 ~ general(-log(s2)); prior llambda: ~ general(lp); lambda = exp(llambda[pump]); model y ~ poisson(lambda); run;
Example 52.5: Random-Effects Models
F
4249
The PROC MCMC statement specifies the input data set ( pump
), the output data set ( postout
), a seed for the random number generator, and an MCMC sample of 100000. It also requests a tuning sample size of 2000 and a thinning rate of 10. The
option keeps track of a number of parameters and symbols in the model. The five parameters are beta1
, beta2
, alpha1
, alpha2
, and s2
.
The symbol logsig is the log of the standard deviation, adif and alpha2
, and bdif measures the difference between displays the summary statistics table.
beta1 measures the difference between alpha1 and beta2
. The ODS SELECT statement
Modeling the random effects to modeling log ij e ij with a normal distribution with mean 0 and variance with a normal distribution with mean
˛ i
C ˇ i
.
log t ij log t /
2 is equivalent and variance
2
.
Here again, the prior distribution on log ij depends on the data set variable logstd
; hence, the construction of the prior must take place before the
statement for log ij
. The symbol lp keeps track of the cumulative log prior density for log ij
.
The symbol lambda is the exponential of the corresponding log ij
, and the
statement gives the response variable y a Poisson likelihood with a mean parameter lambda
.
The posterior summary statistics table is shown in
Output 52.5.4
Summary Statistics for the Nonlinear Poisson Regression
Parameter logsig beta1 beta2 alpha1 alpha2 s2 adif bdif
Nonlinear Poisson Regression Random Effects Model
The MCMC Procedure
N
10000
10000
10000
10000
10000
10000
10000
10000
Posterior Summaries
Mean
Standard
Deviation
0.1045
-0.4467
0.5858
2.9719
1.6406
1.7004
1.3313
-1.0325
0.3862
1.2818
0.5808
2.3658
0.8674
1.7995
2.4934
1.4186
25%
Percentiles
50%
-0.1563
-1.1641
0.2376
1.6225
1.1429
0.7316
-0.1670
-1.8379
0.0883
-0.4421
0.5796
2.9612
1.6782
1.1931
1.2669
-1.0284
75%
0.3496
0.2832
0.9385
4.3137
2.1673
2.0121
2.8032
-0.2189
ˇ
1
D
( 1996 ) reports a posterior mean and standard deviation as follows: log
. 0:45; 1:5/ , ˇ
2
D
.0:63; 0:68/ , and ˛1 ˛2
D
D .0:28; 0:42/
.1:3; 3:0/ . Most estimates from
, agree with Draper’s estimates, with the exception of log different set of prior distributions on
˛ i
,
ˇ i
, and
. The difference might be attributed to the that are used in this analysis.
4250
F
Chapter 52: The MCMC Procedure
You can also use PROC NLMIXED to fit the same model. The following statements run PROC
NLMIXED and produce
proc nlmixed data=pump; ods select parameterestimates additionalestimates; ods output additionalestimates=cp parameterestimates=ps; parms logsig 0 beta1 1 beta2 1 alpha1 1 alpha2 1; if (group = 1) then eta = alpha1 + beta1*logtstd + e; else eta = alpha2 + beta2*logtstd + e; lambda = exp(eta); model y ~ poisson(lambda); random e ~ normal(0,exp(2*logsig)) subject=pump; estimate 'adif' alpha1-alpha2; estimate 'bdif' beta1-beta2; estimate 's2' exp(2*logsig); run;
Output 52.5.5
Estimates by PROC NLMIXED
Parameter logsig beta1 beta2 alpha1 alpha2
Nonlinear Poisson Regression Random Effects Model
The NLMIXED Procedure
Estimate
-0.3161
-0.4256
0.6097
2.9644
1.7992
Parameter Estimates
Standard
Error DF t Value Pr > |t|
0.3213
0.7473
0.3814
1.3826
0.5492
9
9
9
9
9
-0.98
-0.57
1.60
2.14
3.28
Parameter Estimates
0.3508
0.5829
0.1443
0.0606
0.0096
Alpha
0.05
0.05
0.05
0.05
0.05
Parameter logsig beta1 beta2 alpha1 alpha2
Upper
0.4107
1.2649
1.4724
6.0921
3.0415
Gradient
-0.00002
-0.00002
-1.61E-6
-5.25E-6
-5.73E-6
Lower
-1.0429
-2.1162
-0.2530
-0.1632
0.5568
Additional Estimates
Label Estimate
Standard
Error adif bdif s2
1.1653
-1.0354
0.5314
1.4855
0.8389
0.3415
DF t Value Pr > |t|
9
9
9
0.78
-1.23
1.56
0.4529
0.2484
0.1541
Alpha
0.05
0.05
0.05
Lower
-2.1952
-2.9331
-0.2410
Upper
4.5257
0.8623
1.3038
Example 52.5: Random-Effects Models
F
4251
Again, the point estimates from PROC NLMIXED for the mean parameters agree relatively closely with the Bayesian posterior means. You can note that there are differences in the likelihood-based standard errors. This is most likely due to the fact that the Bayesian standard deviations account for the uncertainty in estimating
2
, whereas the likelihood approach plugs in its estimated value.
You can do a similar kernel density plot that compares the PROC MCMC results, the PROC
NLMIXED results and those reported by Draper. The following statements generate
data nlmest(keep = parm mean sd); set ps cp; mean = estimate; sd = standarderror; if _n_ <= 5 then parm = parameter; else parm = label; run; data msd (keep=mean sd); set nlmest; do j = 1 to 401; output; end; run; data _null_; set ps; call symputx(compress('m' || parameter,'*'), estimate); call symputx(compress('s' || parameter,'*'), standarderror); run; data _null_; set cp; call symputx(compress('m' || label,'*'), estimate); call symputx(compress('s' || label,'*'), standarderror); run;
%put &mlogsig &mbeta1 &mbeta2 &malpha1 &malpha2 &madif &mbdif &ms2;
%put &slogsig &sbeta1 &sbeta2 &salpha1 &salpha2 &sadif &sbdif &ss2; proc kde data=postout; univar logsig beta1 beta2 alpha1 alpha2 adif bdif s2 / out=m1 (drop=count); run;
%reshape(m1, mcmc, suffix1=, suffix2=_kde); data all; set mcmc; logsig_nlm = pdf('normal', logsig, &mlogsig, &slogsig); alpha1_nlm = pdf('normal', alpha1, &malpha1, &salpha1); alpha2_nlm = pdf('normal', alpha2, &malpha2, &salpha2); beta1_nlm = pdf('normal', beta1, &mbeta1, &sbeta1); beta2_nlm = pdf('normal', beta2, &mbeta2, &sbeta2);
4252
F
Chapter 52: The MCMC Procedure
adif_nlm = pdf('normal', adif, &madif, &sadif); bdif_nlm = pdf('normal', bdif, &mbdif, &sbdif); s2_nlm = pdf('normal', s2, &ms2, &ss2); logsig_draper = pdf('normal', logsig, 0.28, 0.42); beta1_draper = pdf('normal', beta1, -0.45, 1.5); beta2_draper = pdf('normal', beta2, 0.63, 0.68); adif_draper = pdf('normal', adif, 1.3, 3.0); run; proc template; define statgraph threebythree;
%macro plot; begingraph; layout lattice / rows=3 columns=3; layout overlay /yaxisopts=(label=" "); seriesplot y=logsig_kde x=logsig
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=logsig_nlm x=logsig
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; seriesplot y=logsig_draper x=logsig
/ connectorder=xaxis lineattrs=(pattern=shortdash color=green) legendlabel = "Draper (1996) Approximation" name="Draper"; endlayout;
%do i = 1 %to 2; layout overlay /yaxisopts=(label=" "); seriesplot y=alpha&i._kde x=alpha&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=alpha&i._nlm x=alpha&i
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; endlayout;
%end;
%do i = 1 %to 2; layout overlay /yaxisopts=(label=" "); seriesplot y=beta&i._kde x=beta&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=beta&i._nlm x=beta&i
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; seriesplot y=beta&i._draper x=beta&i
/ connectorder=xaxis lineattrs=(pattern=shortdash color=green) legendlabel = "Draper (1996) Approximation" name="Draper"; endlayout;
%end; layout overlay /yaxisopts=(label=" ");
Example 52.5: Random-Effects Models
F
4253
seriesplot y=adif_kde x=adif
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=adif_nlm x=adif
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; seriesplot y=adif_draper x=adif
/ connectorder=xaxis lineattrs=(pattern=shortdash color=green) legendlabel = "Draper (1996) Approximation" name="Draper"; endlayout; layout overlay /yaxisopts=(label=" "); seriesplot y=bdif_kde x=bdif
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=bdif_nlm x=bdif
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; endlayout; layout overlay /yaxisopts=(label=" ") xaxisopts=(linearopts=(viewmin=0 viewmax=5)); seriesplot y=s2_kde x=s2
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC Kernel" name="MCMC"; seriesplot y=s2_nlm x=s2
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "NLMIXED Approximation" name="NLMIXED"; endlayout;
Sidebar / align = bottom; discretelegend "MCMC" "NLMIXED" "Draper"; endsidebar; endlayout; endgraph;
%mend; %plot; end; run; proc sgrender data=all template=threebythree; run;
The macro %RESHAPE is defined in the example “ Logistic Regression Random-Effects Model ” on
page 4238.
4254
F
Chapter 52: The MCMC Procedure
Output 52.5.6
Comparing Estimates from PROC MCMC (dashed blue), PROC NLMIXED (solid red) and Draper (dotted green)
Consider the data set from
stagnant surface layer and the covariate x i
y i is the logarithm of the height of the is the logarithm of the flow rate of water. The following statements create the data set:
title 'Change Point Model'; data stagnant; input y x @@; ind = _n_; datalines;
1.12
-1.39
0.92
0.65
-0.94
-0.25
1.12
0.90
0.67
0.44
-1.39
-0.80
-0.25
0.11
0.51
0.33
0.13
-0.30
-0.65
;
0.01
0.25
0.44
0.85
1.19
0.30
-0.01
-0.33
0.25
0.59
0.85
0.99
0.81
0.60
0.43
0.25
-0.13
-0.46
-1.08
-0.63
-0.12
0.11
0.34
0.70
0.99
1.03
0.83
0.59
0.43
0.24
-0.14
-0.43
-1.08
-0.63
-0.12
0.11
0.34
0.70
0.99
Example 52.6: Change Point Models
F
4255
A scatter plot ( Output 52.6.1
) shows the presence of a nonconstant slope in the data. This suggests a
change point regression model ( Carlin, Gelfand, and Smith 1992 ). The following statements generate
the scatter plot in
proc sgplot data=stagnant; scatter x=x y=y; run;
Output 52.6.1
Scatter Plot of the Stagnant Data Set
Let the change point be cp
. Following formulation by
model is as follows: y i normal normal
.˛
.˛
C
C
ˇ
ˇ
1
2
.x
.x
i i cp
/;
2
/ cp
/;
2
/ if if x x i i
< cp
>
D cp
You might consider the following diffuse prior distributions:
.
cp
/
.˛; ˇ
1
; ˇ
2
/
.
2
/ uniform
. 1:3; 1:1/ normal .0; var
D
1e6/ uniform .0; 5/
The following statements generate
proc mcmc data=stagnant outpost=postout seed=24860 ntu=1000 nmc=20000;
4256
F
Chapter 52: The MCMC Procedure
ods select PostSummaries; ods output PostSummaries=ds; array beta[2]; parms alpha cp beta1 beta2; parms s2; prior cp ~ unif(-1.3, 1.1); prior s2 ~ uniform(0, 5); prior alpha beta: ~ normal(0, v = 1e6); j = 1 + (x >= cp); mu = alpha + beta[j] * (x - cp); model y ~ normal(mu, var=s2); run;
The PROC MCMC statement specifies the input data set ( stagnant ), the output data set ( postout ), a random number seed, a tuning sample of 1000, and an MCMC sample of 20000. The ODS
SELECT statement displays only the summary statistics table. The ODS OUTPUT statement saves the summary statistics table to the data set ds .
The
statement allocates an array of size 2 for the beta parameters. You can use beta1 and beta2 as parameter names without allocating an array, but having the array makes it easier to construct the likelihood function. The two
statements put the five model parameters in two blocks.
The three
statements specify the prior distributions for these parameters.
The symbol point, j indicates the segment component of the regression. When
(x >= cp) returns 0 and j is assigned the value 1; if x x is less than the change is greater than or equal to the change point,
(x >= cp) returns 1 and j is 2. The symbol mu is the mean for the j th segment, and beta[j] changes between the two regression coefficients depending on the segment component. The
statement assigns the normal model to the response variable y
.
Posterior summary statistics are shown in
Output 52.6.2
MCMC Estimates of the Change Point Regression Model
Parameter alpha cp beta1 beta2 s2
N
20000
20000
20000
20000
20000
Change Point Model
The MCMC Procedure
Posterior Summaries
Mean
0.5349
0.0283
-0.4200
-1.0136
0.000451
Standard
Deviation
0.0249
0.0314
0.0146
0.0167
0.000145
25%
Percentiles
50%
0.5188
0.00728
-0.4293
-1.0248
0.000348
0.5341
0.0303
-0.4198
-1.0136
0.000425
75%
0.5509
0.0493
-0.4111
-1.0023
0.000522
Example 52.6: Change Point Models
F
4257
You can use PROC SGPLOT to visualize the model fit.
shows the fitted regression lines over the original data. In addition, on the bottom of the plot is the kernel density of the posterior marginal distribution of cp
, the change point. The kernel density plot shows the relative variability of the posterior distribution on the data plot. You can use the following statements to create the plot:
data _null_; set ds; call symputx(parameter, mean); run; data b; missing A; input x1 @@; if x1 eq .A then x1 = &cp; if _n_ <= 2 then y1 = &alpha + &beta1 * (x1 - &cp); else y1 = &alpha + &beta2 * (x1 - &cp); datalines;
-1.5 A 1.2
; proc kde data=postout; univar cp / out=m1 (drop=count); run; data m1; set m1; density = (density / 25) - 0.653; run; data all; set stagnant b m1; run; proc sgplot data=all noautolegend; scatter x=x y=y; series x=x1 y=y1 / lineattrs = graphdata2; series x=value y=density / lineattrs = graphdata1; run;
The macro variables data set ds
&alpha
. The data set
,
&beta1
,
&beta2
, and
&cp store the posterior mean estimates from the predicted contains three predicted values, at the minimum and maximum values of x and the estimated change point
&cp
. These input values give you fitted values from the regression model. Data set m1 contains the kernel density estimates of the parameter cp
. The density is scaled down so the curve would fit in the plot. Finally, you use PROC SGPLOT to overlay the scatter plot, regression line and kernel density plots in the same graph.
4258
F
Chapter 52: The MCMC Procedure
Output 52.6.3
Estimated Fit to the Stagnant Data Set
This example covers two commonly used survival analysis models: the exponential model and the Weibull model. The deviance information criterion (DIC) is used to do model selections, and you can also find programs that visualize posterior quantities. Exponential and Weibull models are widely used for survival analysis. This example shows you how to use PROC MCMC to analyze the treatment effect for the E1684 melanoma clinical trial data. These data were collected to assess the effectiveness of using interferon alpha-2b in chemotherapeutic treatment of melanoma. The following statements create the data set:
data e1684; input t t_cen treatment @@; if t = . then do; t = t_cen; v = 0; end; else v = 1; ifn = treatment - 1; et = exp(t); lt = log(t); drop t_cen;
Example 52.7: Exponential and Weibull Survival Analysis
F
4259
1.57808
2.23288
.
datalines;
0.00000
0.00000
9.64384
2
1
1
... more lines ...
1.48219
.
1.66575
0.00000
9.38356
0.00000
2
2
2
.
3.27671
0.94247
7.33425
0.00000
0.00000
1
1
1
;
3.39178
0.00000
1 .
4.36164
2 .
4.81918
2
The data set e1684 contains the following variables: t is the failure time that equals the censoring time whether the observation was censored, v indicates whether the observation is an actual failure time or a censoring time, treatment indicates two levels of treatments, and ifn indicates the use of interferon as a treatment. The variables et and lt are the exponential and logarithm transformation of the time t
. The published data contains other potential covariates that are not listed here. This example concentrates on the effectiveness of the interferon treatment.
Exponential Survival Model
The density function for exponentially distributed survival times is as follows: p.t
i j i
/
D i exp . i t i
/
Note that this formulation of the exponential distribution is different from what is used in the SAS probability function PDF. The definition used in PDF for the exponential distributions is as follows: p.t
i j i
/ D
1 i exp
.
t i i
/ and is as follows: The relationship between i
D
1 i
The corresponding survival function, using the i formulation, is as follows:
S.t
i j i
/ D exp
. i t i
/
If you have a sample f t i g of n likelihood function in terms of independent exponential survival times, each with mean is as follows: i
, then the
L.
j t / D
D
D
… n
… i D 1 n i D 1
… n i D 1 p.t
i j
.
i i i exp i exp
/ i
S.t
.
. i i t t i i i
/ j
// i
/
1 i i .
exp
. i t i
//
1 i
If you link the covariates to to the i th observation and
ˇ with i
D exp x i
0
ˇ , where x i is the vector of covariates corresponding is a vector of regression coefficients, then the log-likelihood function is as follows: l.ˇ j t; x/ D n
X i D 1 i x i
0
ˇ t i exp
.x
i
0
ˇ/
4260
F
Chapter 52: The MCMC Procedure
In the absence of prior information about the parameters in this model, you can choose diffuse normal priors for the
ˇ
:
ˇ normal .0; sd = 10000/
There are two ways to program the log-likelihood function in PROC MCMC. You can use the SAS functions LOGPDF and LOGSDF. Alternatively, you can use the simplified log-likelihood function, which is more computationally efficient. You get identical results by using either approaches.
The following PROC MCMC statements fit an exponential model with simplified log-likelihood function:
title 'Exponential Survival Model'; ods graphics on; proc mcmc data=e1684 outpost=expsurvout nmc=10000 seed=4861; ods select PostSummaries PostIntervals TADpanel ess mcse; parms (beta0 beta1) 0; prior beta: ~ normal(0, sd = 10000);
/*****************************************************/
/* (1) the logpdf and logsdf functions are not used
*/
/*****************************************************/
/* nu = 1/exp(beta0 + beta1*ifn); llike = v*logpdf("exponential", t, nu) +
(1-v)*logsdf("exponential", t, nu);
*/
/****************************************************/
/* (2) the simplified likelihood formula is used
*/
/****************************************************/ l_h = beta0 + beta1*ifn; llike = v*(l_h) t*exp(l_h); model general(llike); run; ods graphics off;
The two assignment statements that are commented out calculate the log-likelihood function by using the SAS functions LOGPDF and LOGSDF for the exponential distribution. The next two assignment statements calculate the log likelihood by using the simplified formula. The first approach is slower because of the redundant calculation involved in calling both LOGPDF and LOGSDF.
An examination of the trace plots for ˇ
0 and ˇ
1
(see
) reveals that the sampling has
gone well with no particular concerns about the convergence or mixing of the chains.
Example 52.7: Exponential and Weibull Survival Analysis
F
4261
Output 52.7.1
Posterior Plots for
ˇ
0 and
ˇ
1 in the Exponential Survival Analysis
The MCMC results are shown in
4262
F
Chapter 52: The MCMC Procedure
Output 52.7.2
Posterior Summary and Interval Statistics
Parameter beta0 beta1
N
10000
10000
Exponential Survival Model
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-1.6715
-0.2879
0.1091
0.1615
25%
Percentiles
50%
-1.7426
-0.4001
-1.6684
-0.2892
75%
-1.5964
-0.1803
Parameter beta0 beta1
Alpha
0.050
0.050
Posterior Intervals
Equal-Tail Interval
-1.8907
-0.5985
-1.4639
0.0300
HPD Interval
-1.8930
-0.6104
-1.4673
0.0169
The Monte Carlo standard errors and effective sample sizes are shown in
means for
ˇ
0 and
ˇ
1 are estimated with high precision, with small standard errors with respect to the standard deviation. This indicates that the mean estimates have stabilized and do not vary greatly in the course of the simulation. The effective sample sizes are roughly the same for both parameters.
Output 52.7.3
MCSE and ESS
Exponential Survival Model
The MCMC Procedure
Parameter
Monte Carlo Standard Errors
MCSE
Standard
Deviation beta0 beta1
0.00302
0.00485
0.1091
0.1615
MCSE/SD
0.0277
0.0301
Parameter beta0 beta1
Effective Sample Sizes
ESS
1304.1
1107.2
Autocorrelation
Time
7.6682
9.0319
Efficiency
0.1304
0.1107
The next part of this example shows fitting a Weibull regression to the data and then comparing the two models with DIC to see which one provides a better fit to the data.
Example 52.7: Exponential and Weibull Survival Analysis
F
4263
Weibull Survival Model
The density function for Weibull distributed survival times is as follows: p.t
i j
˛; i
/
D
˛t i
˛ 1 exp .
i exp .
i
/t i
˛
/
Note that this formulation of the Weibull distribution is different from what is used in the SAS probability function PDF. The definition used in PDF is as follows: p.t
i j ˛; i
/ D exp t i i
˛
˛ i t i i
˛ 1
The relationship between and in these two parameterizations is as follows: i
D ˛ log i
The corresponding survival function, using the i formulation, is as follows:
S.t
i j ˛; i
/ D exp
.
exp
.
i
/t i
˛
/
If you have a sample f t i g of n independent Weibull survival times, with parameters the likelihood function in terms of
˛ and is as follows:
˛
, and i
, then
L.˛; j t / D
D
D
… n i D 1 p.t
i j ˛; i
/
… n i D 1 n
… i D 1
.˛t i
˛ 1
.˛t i
˛ 1 exp exp i S.t
.
.
i i
// i j i
˛; exp
.
i
/
.
exp .
1 i i
/t i
˛
// exp i
.
.
i exp
/t i
˛
.
// exp .
i
/t i
˛
//
1 i
If you link the covariates to the i th observation and
ˇ with i
D x i
0
ˇ
, where x i is the vector of covariates corresponding to is a vector of regression coefficients, the log-likelihood function becomes this: l.˛; ˇ j t; x/ D n
X i D 1 i
.
log
.˛/ C .˛ 1/ log
.t
i
/ C x i
0
ˇ/ exp
.x
i
0
ˇ/t i
˛
/
As with the exponential model, in the absence of prior information about the parameters in this model, you can use diffuse normal priors on ˇ: You might wish to choose a diffuse gamma distribution for
˛:
Note that when
˛ D 1
, the Weibull survival likelihood reduces to the exponential survival likelihood. Equivalently, by looking at the posterior distribution of
˛
, you can conclude whether fitting an exponential survival model would be more appropriate than the Weibull model.
PROC MCMC also allows you to make inference on any functions of the parameters. Quantities of interest in survival analysis include the value of the survival function at specific times for specific treatments and the relationship between the survival curves for different treatments. With PROC
MCMC, you can compute a sample from the posterior distribution of the interested survival functions at any number of points. The data in this example range from about 0 to 10 years, and the treatment of interest is the use of interferon.
Like in the previous exponential model example, there are two ways to fit this model: using the SAS functions LOGPDF and LOGSDF, or using the simplified log likelihood functions. The example uses the latter method. The following statements run PROC MCMC and produce
4264
F
Chapter 52: The MCMC Procedure
title 'Weibull Survival Model'; proc mcmc data=e1684 outpost=weisurvout nmc=10000 seed=1234 monitor=(_parms_ surv_ifn surv_noifn); ods select PostSummaries; ods output PostSummaries=ds PostIntervals=is; array surv_ifn[10]; array surv_noifn[10]; parms alpha 1 (beta0 beta1) 0; prior beta: ~ normal(0, var=10000); prior alpha ~ gamma(0.001,is=0.001); beginnodata; do t = 1 to 10; surv_ifn[t] = exp(-exp(beta0+beta1)*t**alpha); surv_noifn[t] = exp(-exp(beta0)*t**alpha); end; endnodata; lambda = beta0 + beta1*ifn;
/*****************************************************/
/* (1) the logpdf and logsdf functions are not used
*/
/*****************************************************/
/* gamma = exp(-lambda /alpha); llike = v*logpdf('weibull', t, alpha, gamma) +
(1-v)*logsdf('weibull', t, alpha, gamma);
*/
/****************************************************/
/* (2) the simplified likelihood formula is used
*/
/****************************************************/ llike = v*(log(alpha) + (alpha-1)*log(t) + lambda) exp(lambda)*(t**alpha); model general(llike); run;
The
option indicates the parameters and quantities of interest that PROC MCMC tracks.
The symbol
_PARMS_ specifies all model parameters. The array surv_ifn stores the expected survival probabilities for patients who received interferon over a period of 10 years. Similarly, surv_noifn stores the expected survival probabilities for patients who did not received interferon
.
The
and
statements enclose the calculations for the survival probabilities. The assignment statements proceeding the
statement calculate the log likelihood for the Weibull survival model. The
statement specifies the log likelihood that you programmed.
An examination of the trace plots for ˛ , ˇ
0
, and ˇ
1
(not displayed here) reveals that the sampling has gone well, with no particular concerns about the convergence or mixing of the chains.
displays the posterior summary statistics.
Example 52.7: Exponential and Weibull Survival Analysis
F
4265
Output 52.7.4
Posterior Summary Statistics
Parameter alpha beta0 beta1 surv_ifn1 surv_ifn2 surv_ifn3 surv_ifn4 surv_ifn5 surv_ifn6 surv_ifn7 surv_ifn8 surv_ifn9 surv_ifn10 surv_noifn1 surv_noifn2 surv_noifn3 surv_noifn4 surv_noifn5 surv_noifn6 surv_noifn7 surv_noifn8 surv_noifn9 surv_noifn10
N
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
10000
Weibull Survival Model
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
0.7856
-1.3414
-0.2918
0.8212
0.7128
0.6283
0.5588
0.5001
0.4497
0.4060
0.3677
0.3340
0.3041
0.7685
0.6360
0.5372
0.4593
0.3960
0.3437
0.2999
0.2629
0.2313
0.2041
0.0533
0.1389
0.1683
0.0237
0.0308
0.0352
0.0383
0.0405
0.0420
0.0431
0.0437
0.0440
0.0440
0.0280
0.0349
0.0386
0.0407
0.0417
0.0421
0.0419
0.0412
0.0403
0.0392
0.7488
-1.4321
-0.4050
0.8054
0.6919
0.6039
0.5326
0.4728
0.4204
0.3760
0.3375
0.3035
0.2736
0.7501
0.6131
0.5119
0.4330
0.3686
0.3159
0.2715
0.2349
0.2037
0.1767
0.7849
-1.3424
-0.2919
0.8228
0.7138
0.6282
0.5589
0.4997
0.4489
0.4051
0.3664
0.3325
0.3024
0.7701
0.6376
0.5384
0.4599
0.3959
0.3432
0.2993
0.2624
0.2302
0.2033
25%
Percentiles
50% 75%
0.8225
-1.2463
-0.1711
0.8374
0.7337
0.6522
0.5852
0.5276
0.4786
0.4350
0.3972
0.3638
0.3337
0.7876
0.6598
0.5637
0.4878
0.4256
0.3727
0.3282
0.2909
0.2581
0.2302
An examination of the
˛ parameter reveals that the exponential model might not be inappropriate here. The estimated posterior mean of ˛ is 0.7856 with a posterior standard deviation of 0.0533.
As noted previously, if
˛ D 1
, then the Weibull survival distribution is the exponential survival distribution. With these data, you can see that the evidence is in favor of
˛ < 1
. The value 1 is almost
4 posterior standard deviations away from the posterior mean. The following statements compute the posterior probability of the hypothesis that
˛ < 1:
:
proc format; value alphafmt low-<1 = 'alpha < 1' 1-high = 'alpha >= 1'; run; proc freq data=weisurvout; tables alpha /nocum; format alpha alphafmt.; run;
The PROC FREQ results are shown in
4266
F
Chapter 52: The MCMC Procedure
Output 52.7.5
Frequency Analysis of
˛
Weibull Survival Model
The FREQ Procedure alpha Frequency Percent
----------------------------------alpha < 1 10000 100.00
The output from PROC FREQ shows that 100% of the 10000 simulated values for
˛ are less than 1.
This is a very strong indication that the exponential model is too restrictive to model these data well.
You can examine the estimated survival probabilities over time individually, either through the posterior summary statistics or by looking at the kernel density plots. Alternatively, you might find it more informative to examine these quantities in relation with each other. For example, you can use a
side-by-side box plot to display these posterior distributions by using PROC SGPLOT (“ Statistical
Graphics Using ODS ” on page 605). First you need to take the posterior output data set
weisurvout and stack variables that you want to plot. For example, to plot all the survival times for patients who received interferon
, you want to stack surv_inf1
– surv_inf10
. The macro
%Stackdata takes an input data set dataset
, stacks the wanted variables vars
, and outputs them into the output data set.
The following statements define the macro stackdata
:
/* define macro stackdata */
%macro StackData(dataset,output,vars); data &output; length var $ 32; if 0 then set &dataset nobs=nnn; array lll[*] &vars; do jjj=1 to dim(lll); do iii=1 to nnn; set &dataset point=iii; value = lll[jjj]; call vname(lll[jjj],var); output; end; end; stop; keep var value; run;
%mend;
/* stack the surv_ifn variables and saved them to survifn. */
%StackData(weisurvout, survifn, surv_ifn1-surv_ifn10);
Once you stack the data, use PROC SGPLOT to create the side-by-side box plots. The following statements generate
proc sgplot data=survifn; yaxis label='Survival Probability' values=(0 to 1 by 0.2); xaxis label='Time' discreteorder=data;
Example 52.7: Exponential and Weibull Survival Analysis
F
4267
vbox value / category=var; run;
Output 52.7.6
Side-by-Side Box Plots of Estimated Survival Probabilities
There is a clear decreasing trend over time of the survival probabilities for patients who receive the treatment. You might ask how does this group compare to those who did not receive the treatment?
In this case, you want to overlay the two predicted curves for the two groups of patients and add the corresponding credible interval. See
. To generate the graph, you first take the posterior
mean estimates from the ODS output table ds and the lower and upper HPD interval estimates store them in the data set surv , and draw the figure by using PROC SGPLOT.
is
,
The following statements generate data set surv :
data surv; set ds; if _n_ >= 4 then do; set is point=_n_; group = 'with interferon time = _n_ - 3; if time > 10 then do; time = time - 10;
'; group = 'without interferon'; end; output; end; keep time group mean hpdlower hpdupper; run;
4268
F
Chapter 52: The MCMC Procedure
The following SGPLOT statements generate
proc sgplot data=surv; yaxis label="Survival Probability" values=(0 to 1 by 0.2); series x=time y=mean / group = group name='i'; band x=time lower=hpdlower upper=hpdupper / group = group transparency=0.7; keylegend 'i'; run;
In
, the solid line is the survival curve for patients who received
interferon
; the shaded region centers at the solid line is the 95% HPD intervals; the medium-dashed line is the survival curve for patients who did not receive interferon
; and the shaded region around the dashed line is the corresponding 95% HPD intervals.
Output 52.7.7
Predicted Survival Probability Curves with 95% HPD Intervals
The plot suggests that there is an effect of using interferon because patients who received interferon have sustained better survival probabilities than those who did not. However, the effect might not be very significant, as the 95% credible intervals of the two groups do overlap. For more on these interferon studies, refer to
Ibrahim, Chen, and Sinha ( 2001 ).
Example 52.7: Exponential and Weibull Survival Analysis
F
4269
Weibull or Exponential?
Although the evidence from the Weibull model fit shows that the posterior distribution of ˛ has a significant amount of density mass less than 1, suggesting that the Weibull model is a better fit to the data than the exponential model, you might still be interested in comparing the two models more
formally. You can use the Bayesian model selection criterion (see the section “ Deviance Information
Criterion (DIC) ” on page 171) to determine which model fits the data better.
The PROC MCMC
option requests the calculation of DIC, and the procedure displays the
ODS output table at the estimate,
D.
DIC . The table includes the posterior mean of the deviation,
/
, effective number of parameters, p
D
D.
/
, deviation
, and DIC. It is important to remember that the standardizing term, p.
y / , which is a function of the data alone, is not taken into account in calculating the DIC. This term is irrelevant only if you compare two models that have the same likelihood function. If you do not have identical likelihood functions, using DIC for model selection purposes without taking this standardizing term into account can produce incorrect results. In addition, you want to be careful in interpreting the DIC whenever you use the
function to construct the log-likelihood, as the case in this example. Using the
function, you can obtain identical posterior samples with two log-likelihood functions that differ only by a constant.
This difference translates to a difference in the DIC calculation, which could be very misleading.
If
˛ D 1
, the Weibull likelihood is identical to the exponential likelihood. It is safe in this case to directly compare DICs from these two models. However, if you do not want to work out the mathematical detail or you are uncertain of the equivalence, a better way of comparing the DICs is to run the Weibull model twice: once with
˛ being a parameter and once with
˛ D the likelihood functions are the same, and the DIC comparison is meaningful.
1
. This ensures that
The following statements fit a Weibull model:
title 'Model Comparison between Weibull and Exponential'; proc mcmc data=e1684 outpost=weisurvout nmc=10000 seed=4861 dic; ods select dic; parms alpha 1 (beta0 beta1) 0; prior beta: ~ normal(0, var=10000); prior alpha ~ gamma(0.001,is=0.001); lambda = beta0 + beta1*ifn; llike = v*(log(alpha) + (alpha-1)*log(t) + lambda) exp(lambda)*(t**alpha); model general(llike); run;
The
option requests the calculation of DIC, and the table is displayed is displayed in
4270
F
Chapter 52: The MCMC Procedure
Output 52.7.8
DIC Table from the Weibull Model
Model Comparison between Weibull and Exponential
The MCMC Procedure
Deviance Information Criterion
Dbar (posterior mean of deviance)
Dmean (deviance evaluated at posterior mean) pD (effective number of parameters)
DIC (smaller is better)
858.623
855.633
2.990
861.614
The GENERAL or DGENERAL function is used in this program.
To make meaningful comparisons, you must ensure that all
GENERAL or DGENERAL functions include appropriate normalizing constants. Otherwise, DIC comparisons can be misleading.
The note in
reminds you of the importance of ensuring identical likelihood functions when you use the
function. The DIC value is 861:6 .
Based on the same set of code, the following statements fit an exponential model by setting
˛ D 1
:
proc mcmc data=e1684 outpost=expsurvout nmc=10000 seed=4861 dic; ods select dic; parms beta0 beta1 0; prior beta: ~ normal(0, var=10000); begincnst; alpha = 1; endcnst; lambda = beta0 + beta1*ifn; llike = v*(log(alpha) + (alpha-1)*log(t) + lambda) exp(lambda)*(t**alpha); model general(llike); run;
displays the DIC table.
Example 52.8: Cox Models
F
4271
Output 52.7.9
DIC Table from the Exponential Model
Model Comparison between Weibull and Exponential
The MCMC Procedure
Deviance Information Criterion
Dbar (posterior mean of deviance)
Dmean (deviance evaluated at posterior mean) pD (effective number of parameters)
DIC (smaller is better)
870.133
868.190
1.943
872.075
The GENERAL or DGENERAL function is used in this program.
To make meaningful comparisons, you must ensure that all
GENERAL or DGENERAL functions include appropriate normalizing constants. Otherwise, DIC comparisons can be misleading.
The DIC value of
872:075 is greater than
861
. A smaller DIC indicates a better fit to the data; hence, you can conclude that the Weibull model is more appropriate for this data set. You can see the
equivalencing of the exponential model you fitted in “ Exponential Survival Model ” on page 4259 by
running the following comparison.
The following statements are taken from the section “ Exponential Survival Model ” on page 4259,
and they fit the same exponential model:
proc mcmc data=e1684 outpost=expsurvout1 nmc=10000 seed=4861 dic; ods select none; parms (beta0 beta1) 0; prior beta: ~ normal(0, sd = 10000); l_h = beta0 + beta1*ifn; llike = v*(l_h) t*exp(l_h); model general(llike); run; proc compare data=expsurvout compare=expsurvout1; var beta0 beta1; run;
The posterior samples of beta0 and beta1 in the data set expsurvout1 are identical to those in the data set expsurvout . The comparison results are not shown here.
This example has two purposes. One is to illustrate how to use PROC MCMC to fit a Cox proportional hazard model. Specifically, two models are considered: time independent and time dependent models.
However, note that it is much easier to fit a Bayesian Cox model by specifying the BAYES statement
4272
F
Chapter 52: The MCMC Procedure
in PROC PHREG (see Chapter 64, “ The PHREG Procedure ”). If you are interested only in fitting a
Cox regression survival model, you should use PROC PHREG.
The second objective of this example is to demonstrate how to model data that are not independent.
That is the case where the likelihood for observation i depends on other observations in the data set. In other words, if you work with a likelihood function that cannot be broken down simply as
L.
y /
D
Q n i
L.y
i
/ , you can use this example for illustrative purposes. By default, PROC MCMC assumes that the programming statements and model specification is intended for a single row of observations in the data set. The Cox model is chosen because the complexity in the data structure requires more elaborate coding.
The Cox proportional hazard model is widely used in the analysis of survival time, failure time, or other duration data to explain the effect of exogenous explanatory variables. The data set used in this example is taken from
Krall, Uthoff, and Harley ( 1975 ), who analyzed data from a study on
myeloma in which researchers treated 65 patients with alkylating agents. Of those patients, 48 died during the study and 17 survived. The following statements generate the data set that is used in this example:
data Myeloma; input Time Vstatus LogBUN HGB Platelet Age LogWBC Frac
LogPBM Protein SCalc; label Time='survival time'
VStatus='0=alive 1=dead'; datalines;
1.25
1.25
2.00
1
1
1
2.2175
1.9395
1.5185
9.4
12.0
9.8
1
1
1
67
38
81
3.6628
3.9868
3.8751
1
1
1
1.9542
1.9542
2.0000
12
20
2
10
18
15
... more lines ...
77.00
;
0 1.0792
14.0
1 60 3.6812
0 0.9542
proc sort data = Myeloma; by descending time; run;
0 12 data _null_; set Myeloma nobs=_n; call symputx('N', _n); stop; run;
The variable
Time represents the survival time in months from diagnosis. The variable
VStatus consists of two values, 0 and 1, indicating whether the patient was alive or dead, respectively, at the end of the study. If the value of
VStatus is 0, the corresponding value of
Time is censored. The variables thought to be related to survival are
LogBUN
(log
.
BUN
/ at diagnosis),
HGB
(hemoglobin at diagnosis),
Platelet
(platelets at diagnosis: 0=abnormal, 1=normal),
Age
(age at diagnosis in years),
LogWBC
(log(WBC) at diagnosis),
Frac
(fractures at diagnosis: 0=none, 1=present),
LogPBM
(log percentage of plasma cells in bone marrow),
Protein
(proteinuria at diagnosis), and
SCalc
(serum calcium at diagnosis). Interest lies in identifying important prognostic factors from these explanatory variables. In addition, there are 65 (
&n
) observations in the data set Myeloma . The likelihood used in
Example 52.8: Cox Models
F
4273 these examples is the Brewslow likelihood:
L.ˇ/ D n
Y
2 d i
Y i D 1
4 j D 1
P l 2 R i
3 v i exp .ˇ
0
Z j
.t
i
// exp .ˇ
0
Z l
.t
i
//
5 where
ˇ is the vector parameters n is the total number of observations in the data set t i is the i th time, which can be either event time or censored time
Z l
.t / is the vector explanatory variables for the l th individual at time t d i is the multiplicity of failures at t i
. If there are no ties in time, d i is 1 for all i
.
R i is the risk set for the greater than or equal to t i i th time t i
, which includes all observations that have survival time v i indicates whether the patient is censored. The value 0 corresponds to censoring. Note that the censored time t i enters the likelihood function only through the formation of the risk set
R i
.
Priors on the coefficients are independent normal priors with very large variance (1e6). Throughout this example, the symbol
S represents the term
P bZ l 2 R i represents the regression term exp
.ˇ
0
Z l
.t
i
//
.
ˇ
0
Z j
.t
i
/ in the likelihood, and the symbol
Time Independent Model
The regression model considered in this example uses the following formula:
ˇ
0
Z j
D ˇ
1 logbun
C ˇ
2 hgb
C ˇ
3 platelet
C ˇ
4 age
C
ˇ
5 logwbc
C ˇ
6 frac
C ˇ
7 logpbm
C ˇ
8 protein
C ˇ
9 scalc
The hard part of coding this in PROC MCMC is the construction of the risk set
R i
.
R i contains all observations that have survival time greater than or equal to t i
. First suppose that there are no ties in time. Sorting the data set by the variable time order. Observation i into descending order gives you
R i
’s risk set consists of all data points j such that j < D i that is in the right in the data set. You can cumulatively increment
S in the SAS statements.
With potential ties in time, at observation i
C
1 i
, you need to know whether any subsequent observations, and so on, have the same survival time as t i
. Suppose that the i th, the i
C
1 th, and the i
C
2 th observations all have the same survival time; all three of them need to be included in the risk set calculation. This means that to calculate the likelihood for some observations, you need to access both the previous and subsequent observations in the data set. There are two ways to do this. One is to use the LAG function; the other is to use the option
4274
F
Chapter 52: The MCMC Procedure
The LAG function returns values from a queue (see
SAS Language Reference: Dictionary
). So for the i th observation, you can use LAG1 to access variables from the previous row in the data set. You want to compare the lag1 value of time with the current time value. Depending on whether the two time values are equal, you can add correction terms in the calculation for the risk set
S
.
The following statements sort the data set by time time on top: into descending order, with the largest survival
title 'Cox Model with Time Independent Covariates'; proc freq data=myeloma; ods select none; tables time / out=freqs; run; proc sort data = freqs; by descending time; run; data myelomaM; set myeloma; ind = _N_; run;
The following statements run PROC MCMC and produce
proc mcmc data=myelomaM outpost=outi nmc=50000 ntu=3000 seed=1; ods select PostSummaries PostIntervals; array beta[9]; parms beta: 0; prior beta: ~ normal(0, var=1e6); bZ = beta1 * LogBUN + beta2 * HGB + beta3 * Platelet
+ beta4 * Age + beta5 * LogWBC + beta6 * Frac + beta7 * LogPBM + beta8 * Protein + beta9 * SCalc; if ind = 1 then do;
S = exp(bZ); l = vstatus * bZ; v = vstatus;
/* first observation end; else if (1 < ind < &N) then do; if (lag1(time) ne time) then do; l = vstatus * bZ; l = l - v * log(S); v = vstatus;
S = S + exp(bZ);
/* correct the loglike value
/* reset v count value end; else do; l = vstatus * bZ;
S = S + exp(bZ); v = v + vstatus; end; end; else do;
/* still a tie
*/
*/
/* add # of nonsensored values */
/* last observation
*/
*/
*/
Example 52.8: Cox Models
F
4275
if (lag1(time) ne time) then do; l = - v * log(S);
S = S + exp(bZ);
/* correct the loglike value l = l + vstatus * (bZ - log(S)); end; else do;
S = S + exp(bZ); l = vstatus * bZ - (v + vstatus) * log(S); end; end; model general(l); run;
*/
The symbol bZ is the regression term, which is independent of the time variable. The symbol ind indexes observation numbers in the data set. The symbol
S keeps track of the risk set term for every observation. The symbol l calculates the log likelihood for each observation. Note that the value of l for observation ind is not necessarily the correct log likelihood value for that observation, especially in cases where the observation ind is in the tied times
. Correction terms are added to subsequent values of l when the time variable becomes different in order to make up the difference. The total sum of l calculated over the entire data set is correct. The symbol v keeps track of the sum of as censored data do not enter the likelihood and need to be taken out.
vstatus
,
You use the function LAG1 to detect if two adjacent time values are different. If they are, you know that the current observation is in a different risk set than the last one. You then need to add a correction term to the log likelihood value of l
. The IF-ELSE statements break the observations into three parts: the first observation, the last observation and everything in the middle.
Output 52.8.1
Summary Statistics on Cox Model with Time Independent Explanatory Variables and Ties in the Survival Time, Using PROC MCMC
Cox Model with Time Independent Covariates
Parameter beta1 beta2 beta3 beta4 beta5 beta6 beta7 beta8 beta9
N
50000
50000
50000
50000
50000
50000
50000
50000
50000
The MCMC Procedure
Posterior Summaries
1.7600
-0.1308
-0.2017
-0.0126
0.3373
0.3992
0.3749
0.0106
0.1272
Mean
Standard
Deviation
0.6441
0.0720
0.5148
0.0194
0.7256
0.4337
0.4861
0.0271
0.1064
1.3275
-0.1799
-0.5505
-0.0257
-0.1318
0.0973
0.0464
-0.00723
0.0579
25%
Percentiles
50%
1.7651
-0.1304
-0.1965
-0.0128
0.3505
0.3864
0.3636
0.0118
0.1300
75%
2.1947
-0.0817
0.1351
0.000641
0.8236
0.6804
0.6989
0.0293
0.1997
4276
F
Chapter 52: The MCMC Procedure
Output 52.8.1
continued
Parameter beta1 beta2 beta3 beta4 beta5 beta6 beta7 beta8 beta9
Alpha
0.050
0.050
0.050
0.050
0.050
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
0.4649
-0.2704
-1.2180
-0.0501
-1.1233
-0.4136
-0.5551
-0.0451
-0.0933
3.0214
0.0114
0.8449
0.0257
1.7232
1.2970
1.3593
0.0618
0.3272
HPD Interval
0.5117
-0.2746
-1.2394
-0.0512
-1.1124
-0.4385
-0.5423
-0.0451
-0.0763
3.0465
0.00524
0.7984
0.0245
1.7291
1.2575
1.3689
0.0616
0.3406
An alternative to using the LAG function is to use the PROC option
JOINTMODEL . With this option,
the log-likelihood function you specify applies not to a single observation but to the entire data
set. See “ Modeling Joint Likelihood ” on page 4181 for details on how to properly use this option.
The basic idea is that you store all necessary data set variables in arrays and use only the arrays to construct the log likelihood of the entire data set. This approach works here because for every observation i , you can use index to access different values of arrays to construct the risk set
S
. To use the
option, you need to do some additional data manipulation. You want to create a stop variable for each observation, which indicates the observation number that should be included in
S for that observation. For example, if observations 4, 5, 6 all have the same survival time, the stop value for all of them is 6.
The following statements generate a new data set myelomaM that contains the stop variable:
data myelomaM; merge myelomaM freqs(drop=percent); by descending time; retain stop; if first.time then do; stop = _n_ + count - 1; end; run;
The following SAS program fits the same Cox model by using the
option:
data a; run; proc mcmc data=a outpost=outa nmc=50000 ntu=3000 seed=1 jointmodel; ods select none; array beta[9]; array data[1] / nosymbols; array timeA[1] / nosymbols; array vstatusA[1] / nosymbols; array stopA[1] / nosymbols; array bZ[&n]; array S[&n]; begincnst;
Example 52.8: Cox Models
F
4277
rc = read_array("myelomam", data, "logbun", "hgb", "platelet",
"age", "logwbc", "frac", "logpbm", "protein", "scalc"); rc = read_array("myelomam", timeA, "time"); rc = read_array("myelomam", vstatusA, "vstatus"); rc = read_array("myelomam", stopA, "stop"); endcnst; parms (beta:) 0; prior beta: ~ normal(0, var=1e6); jl = 0;
/* calculate each bZ and cumulatively adding S as if there are no ties.*/ call mult(data, beta, bZ);
S[1] = exp(bZ[1]); do i = 2 to &n;
S[i] = S[i-1] + exp(bZ[i]); end; do i = 1 to &n;
/* correct the S[i] term, when needed. */ if(stopA[i] > i) then do; do j = (i+1) to stopA[i];
S[i] = S[i] + exp(bZ[j]); end; end; jl = jl + vstatusA[i] * (bZ[i] - log(S[i])); end; model general(jl); run; ods select all;
No output tables were produced because this PROC MCMC run produces identical posterior samples as does the previous example.
Because the
option is specified here, you do not need to specify myelomaM as the input data set. An empty data set a is used to speed up the procedure run.
Multiple
statements allocate array symbols that are used to store the parameters ( beta
), the response and the covariates ( data, timeA, vstatusA, and stopA
), and the work space ( bZ and
S
). The data, timeA, vstatusA, and stopA arrays are declared with the /NOSYMBOLS option. This option enables
PROC MCMC to dynamically resize these arrays to match the dimensions of the input data set. See
the section “ READ_ARRAY Function ” on page 4134. The
bZ and
S arrays store the regression term and the risk set term for every observation.
The
and
statements enclose programming statements that read the data set variables into these arrays. The rest of the programming statements construct the log likelihood for the entire data set.
The CALL MULT function calculates the regression term in the model and stores the result in the array bZ
. In the first DO loop, you sum the risk set term
S as if there are no ties in time. This underevaluates some of the
S elements. For observations that have a tied time
, you make the necessary correction to the corresponding
S values. The correction takes place in the second DO loop. Any observation that has a tied time also has a stopA[i] that is different from i
. You add the right terms to
S
4278
F
Chapter 52: The MCMC Procedure and sum up the joint log likelihood jl
. The
statement specifies that the log likelihood takes on the value of jl
.
To see that you get identical results from these two approaches, use PROC COMPARE to compare the posterior samples from two runs:
proc compare data=outi compare=outa; ods select comparesummary; var beta1-beta9; run;
The output is not shown here.
Generally, the
option can be slightly faster than using the default setup. The savings come from avoiding the overhead cost of accessing the data set repeatedly at every iteration. However, the speed gain is not guaranteed because it largely depends on the efficiency of your programs.
PROC PHREG fits the same model, and you get very similar results to PROC MCMC. The following statements run PROC PHREG and produce
proc phreg data=Myeloma; ods select PostSummaries PostIntervals; ods output posteriorsample = phout; model Time*VStatus(0)=LogBUN HGB Platelet Age LogWBC
Frac LogPBM Protein Scalc; bayes seed=1 nmc=10000; run;
Output 52.8.2
Summary Statistics for Cox Model with Time Independent Explanatory Variables and Ties in the Survival Time, Using PROC PHREG
Cox Model with Time Independent Covariates
Parameter
LogBUN
HGB
Platelet
Age
LogWBC
Frac
LogPBM
Protein
SCalc
N
10000
10000
10000
10000
10000
10000
10000
10000
10000
The PHREG Procedure
Bayesian Analysis
1.7610
-0.1279
-0.2179
-0.0130
0.3150
0.3766
0.3792
0.0102
0.1248
Posterior Summaries
Mean
Standard
Deviation
0.6593
0.0727
0.5169
0.0199
0.7451
0.4152
0.4909
0.0267
0.1062
1.3173
-0.1767
-0.5659
-0.0264
-0.1718
0.0881
0.0405
-0.00745
0.0545
25%
Percentiles
50%
1.7686
-0.1287
-0.2360
-0.0131
0.3321
0.3615
0.3766
0.0106
0.1273
75%
2.2109
-0.0789
0.1272
0.000492
0.8253
0.6471
0.7023
0.0283
0.1985
Example 52.8: Cox Models
F
4279
Output 52.8.2
continued
Parameter
LogBUN
HGB
Platelet
Age
LogWBC
Frac
LogPBM
Protein
SCalc
Alpha
0.050
0.050
0.050
0.050
0.050
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
0.4418
-0.2718
-1.1952
-0.0514
-1.2058
-0.3995
-0.5652
-0.0437
-0.0935
3.0477
0.0150
0.8296
0.0259
1.7228
1.2316
1.3671
0.0611
0.3264
HPD Interval
0.4107
-0.2801
-1.1871
-0.0519
-1.1783
-0.4273
-0.5939
-0.0405
-0.0846
2.9958
0.00599
0.8341
0.0251
1.7483
1.2021
1.3241
0.0637
0.3322
shows kernel density plots that compare the posterior marginal distributions of all the beta parameters from the PROC MCMC run and the PROC PHREG run. The following statements generate the comparison:
proc kde data=outi; ods exclude all; univar beta1 beta2 beta3 beta4 beta5 beta6 beta7 beta8 beta9
/ out=m1 (drop=count); run; ods exclude none;
%reshape(m1, mcmc, suffix1=, suffix2=md); data phout; set phout(drop = LogPost Iteration); beta1 = LogBUN; beta4 = Age; beta2 = HGB; beta5 = LogWBC; beta3 = Platelet; beta6 = Frac; beta7 = LogPBM; beta8 = Protein; beta9 = SCalc; drop LogBUN HGB Platelet Age LogWBC Frac LogPBM Protein SCalc; run; proc kde data=phout; ods exclude all; univar beta1 beta2 beta3 beta4 beta5 beta6 beta7 beta8 beta9
/ out=m2 (drop=count); run; ods exclude none;
%reshape(m2, phreg, suffix1=p, suffix2=pd); data all; merge mcmc phreg; run; proc template; define statgraph threebythree;
%macro plot;
4280
F
Chapter 52: The MCMC Procedure
begingraph; layout lattice / rows=3 columns=3;
%do i = 1 %to 9; layout overlay /yaxisopts=(label=" "); seriesplot y=beta&i.md x=beta&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC" name="MCMC"; seriesplot y=beta&i.pd x=beta&i.p
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "PHREG" name="PHREG"; endlayout;
%end;
Sidebar / align = bottom; discretelegend "MCMC" "PHREG"; endsidebar; endlayout; endgraph;
%mend; %plot; end; run; proc sgrender data=all template=threebythree; title "Kernel Density Comparison"; run;
The macro %RESHAPE is defined in the example “ Logistic Regression Random-Effects Model ” on
page 4238. The posterior densities are almost identical to one another.
Example 52.8: Cox Models
F
4281
Output 52.8.3
Comparing Estimates from PROC MCMC and PROC PHREG
Time Dependent Model
To model
Z i
.t
i
/ as a function of the survival time, you can relate time t i formula: to covariates by using this
ˇ
0
Z j
.t
i
/ D .ˇ
1
C ˇ
2 t i
/ logbun
C .ˇ
3
C ˇ
4 t i
/ hgb
C .ˇ
5
C ˇ
6 t i
/ platelet
For illustrational purposes, only three explanatory variables,
LOGBUN
,
HBG
, and
PLATELET
, are used in this example.
Since
Z j
.t
i
/ current time t depends on i t i
, every term in the summation of
P l 2
R i exp
.ˇ
0
Z and all observations that are in the risk set. You can use the l
.t
i
// is a product of the
option, as in the last example, or you can modify the input data set such that every row contains not only the current observation but also all observations that are in the corresponding risk set. When you construct the log likelihood for each observation, you have all the relevant data at your disposal.
The following statements illustrate how you can create a new data set with different risk sets at different rows:
title 'Cox Model with Time Dependent Covariates'; ods select none; proc freq data=myeloma; tables time / out=freqs; run;
4282
F
Chapter 52: The MCMC Procedure
ods select all; proc sort data = freqs; by descending time; run; data myelomaM; set myeloma; ind = _N_; run; data myelomaM; merge myelomaM freqs(drop=percent); by descending time; retain stop; if first.time then do; stop = _n_ + count - 1; end; run;
%macro array(list);
%global mcmcarray;
%let mcmcarray = ;
%do i = 1 %to 32000;
%let v = %scan(&list, &i, %str( ));
%if %nrbquote(&v) ne %then %do; array _&v[&n];
%let mcmcarray = &mcmcarray array _&v[&n] _&v.1 - _&v.&n%str(;); do i = 1 to stop; set myelomaM(keep=&v) point=i;
_&v[i] = &v; end;
%end;
%else %let i = 32001;
%end;
%mend; data z; set myelomaM;
%array(logbun hgb platelet); drop vstatus logbun hgb platelet count stop; run; data myelomaM; merge myelomaM z; by descending time; run;
The data set myelomaM contains 65 observations and 209 variables. For each observation, you see added variables stop
,
_logbun1 through
_logbun65
,
_hgb1 through
_hgb65
, and
_platelet1 through
_platelet65
. The variable stop indicates the number of observations that are in the risk set of the current observation. The rest are transposed values of model covariates of the entire data set. The data set contains a number of missing values. This is due to the fact that only the relevant observations are kept, such as
_logbun
1 to
_logbun stop . The rest of the cells are filled in with missing values. For example, the first observation has a unique survival time of 92 and stop is 1, making it a risk set of itself. You see nonmissing values only in
_logbun1
,
_hgb1
, and
_platelet1
.
Example 52.8: Cox Models
F
4283
The following statements fit the Cox model by using PROC MCMC:
proc mcmc data=myelomaM outpost=outi nmc=50000 ntu=3000 seed=17 missing=ac; ods select PostSummaries PostIntervals; array beta[6];
&mcmcarray parms (beta:) 0; prior beta: ~ normal(0, prec=1e-6); b = (beta1 + beta2 * time) * logbun +
(beta3 + beta4 * time) * hgb +
(beta5 + beta6 * time) * platelet;
S = 0; do i = 1 to stop;
S = S + exp( (beta1 + beta2 * time) * _logbun[i] +
(beta3 + beta4 * time) * _hgb[i] +
(beta5 + beta6 * time) * _platelet[i]); end; loglike = vstatus * (b - log(S)); model general(loglike); run;
Note that the option
is set to
AC
. This is due to missing cells in the input data set. You must use this option so that PROC MCMC retains observations that contain missing values.
The macro variable
&mcmcarray is defined in the earlier part in this example. You can use a
%put
statement to print its value:
%put &mcmcarray;
This statement prints the following:
array _logbun[65] _logbun1 - _logbun65; array _hgb[65] _hgb1 - _hgb65; array
_platelet[65] _platelet1 - _platelet65;
The macro uses the
statement to allocate three arrays, each of which links their corresponding data set variables. This makes it easier to reference these data set variables in the program. The
statement puts all the parameters in the same block. The
statement gives them normal priors with large variance. The symbol b is the regression term, and
S is cumulatively added from 1 to stop for each observation in the DO loop. The symbol loglike completes the construction of log likelihood for each observation and the
statement completes the model specification.
Posterior summary and interval statistics are shown in
4284
F
Chapter 52: The MCMC Procedure
Output 52.8.4
Summary Statistics on Cox Model with Time Dependent Explanatory Variables and
Ties in the Survival Time, Using PROC MCMC
Cox Model with Time Dependent Covariates
Parameter beta1 beta2 beta3 beta4 beta5 beta6
N
50000
50000
50000
50000
50000
50000
The MCMC Procedure
Posterior Summaries
Mean
3.2397
-0.1411
-0.0369
-0.00409
0.3548
-0.0417
Standard
Deviation
0.8226
0.0471
0.1017
0.00360
0.7359
0.0359
2.6835
-0.1722
-0.1064
-0.00656
-0.1634
-0.0661
25%
Percentiles
50%
3.2413
-0.1406
-0.0373
-0.00408
0.3530
-0.0423
75%
3.7830
-0.1092
0.0315
-0.00167
0.8445
-0.0181
Parameter beta1 beta2 beta3 beta4 beta5 beta6
Alpha
0.050
0.050
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
1.6399
-0.2351
-0.2337
-0.0111
-1.0317
-0.1107
4.8667
-0.0509
0.1642
0.00282
1.8202
0.0295
HPD Interval
1.6664
-0.2294
-0.2272
-0.0112
-1.0394
-0.1122
4.8752
-0.0458
0.1685
0.00264
1.8100
0.0269
You can also use the option
to get the same inference and avoid transposing the data for every observation:
proc mcmc data=myelomaM outpost=outa nmc=50000 ntu=3000 seed=17 jointmodel; ods select none; array beta[6]; array logbunA[&n]; array stopA[&n]; array timeA[&n]; array hgbA[&n]; array bZ[&n]; array vstatusA[&n]; array plateletA[&n]; array S[&n]; begincnst; timeA[ind]=time; logbunA[ind]=logbun; plateletA[ind]=platelet; endcnst; vstatusA[ind]=vstatus; hgbA[ind]=hgb; stopA[ind]=stop; parms (beta:) 0; prior beta: ~ normal(0, prec=1e-6); jl = 0; do i = 1 to &n; v1 = beta1 + beta2 * timeA[i]; v2 = beta3 + beta4 * timeA[i]; v3 = beta5 + beta6 * timeA[i]; bZ[i] = v1 * logbunA[i] + v2 * hgbA[i] + v3 * plateletA[i];
Example 52.8: Cox Models
F
4285
/* sum over risk set without considering ties in time. */
S[i] = exp(bZ[i]); if (i > 1) then do; do j = 1 to (i-1); b1 = v1 * logbunA[j] + v2 * hgbA[j] + v3 * plateletA[j];
S[i] = S[i] + exp(b1); end; end; end;
/* make correction to the risk set due to ties in time. */ do i = 1 to &n; if(stopA[i] > i) then do; v1 = beta1 + beta2 * timeA[i]; v2 = beta3 + beta4 * timeA[i]; v3 = beta5 + beta6 * timeA[i]; do j = (i+1) to stopA[i]; b1 = v1 * logbunA[j] + v2 * hgbA[j] + v3 * plateletA[j];
S[i] = S[i] + exp(b1); end; end; jl = jl + vstatusA[i] * (bZ[i] - log(S[i])); end; model general(jl); run;
The multiple
statements allocate array symbols that are used to store the parameters ( beta
), the response ( timeA
), the covariates ( vstatusA, logbunA, hgbA, plateletA, and stopA
), and work space
( bZ and
S
). The bZ and
S arrays store the regression term and the risk set term for every observation.
Programming statements in the
and
statements input the response and covariates from the data set to the arrays.
Using the same technique shown in the example “ Time Independent Model ” on page 4273, the next
DO loop calculates the regression term and corresponding
S for every observation, pretending that there are no ties in time. This means that the risk set for observation i i
. The correction terms are added to the corresponding
S[i] involves only observation
1 to in the second DO loop, conditional on whether the stop variable is greater than the observation count itself. The symbol jl cumulatively adds the log likelihood for the entire data set, and the function.
statement specifies the joint log-likelihood
The following statements run PROC COMPARE and show that the output data set outa contains identical posterior samples as outi
:
proc compare data=outi compare=outa; ods select comparesummary; var beta1-beta6; run;
The results are not shown here.
The following statements use PROC PHREG to fit the same time dependent Cox model:
4286
F
Chapter 52: The MCMC Procedure
proc phreg data=Myeloma; ods select PostSummaries PostIntervals; ods output posteriorsample = phout; model Time*VStatus(0)=LogBUN z2 hgb z3 platelet z4; z2 = Time*logbun; z3 = Time*hgb; z4 = Time*platelet; bayes seed=1 nmc=10000; run;
Coding is simpler than PROC MCMC. See
statistics: for posterior summary and interval
Output 52.8.5
Summary Statistics on Cox Model with Time Dependent Explanatory Variables and
Ties in the Survival Time, Using PROC PHREG
Cox Model with Time Dependent Covariates
The PHREG Procedure
Parameter
LogBUN z2
HGB z3
Platelet z4
Bayesian Analysis
Posterior Summaries
N
10000
10000
10000
10000
10000
10000
Mean
3.2423
-0.1401
-0.0382
-0.00407
0.3778
-0.0419
Standard
Deviation
0.8311
0.0482
0.1009
0.00363
0.7524
0.0364
2.6838
-0.1723
-0.1067
-0.00652
-0.1500
-0.0660
25%
Percentiles
50%
3.2445
-0.1391
-0.0385
-0.00404
0.3389
-0.0425
75%
3.7929
-0.1069
0.0297
-0.00162
0.8701
-0.0178
Parameter
LogBUN z2
HGB z3
Platelet z4
Alpha
0.050
0.050
0.050
0.050
0.050
0.050
Posterior Intervals
Equal-Tail Interval
1.6059
-0.2361
-0.2343
-0.0113
-0.9966
-0.1124
4.8785
-0.0494
0.1598
0.00297
1.9464
0.0296
HPD Interval
1.5925
-0.2354
-0.2331
-0.0109
-1.1342
-0.1142
4.8582
-0.0492
0.1603
0.00322
1.7968
0.0274
shows a kernel density comparison plot that compares posterior marginal distributions of all the beta parameters from the PROC MCMC run and the PROC PHREG run. The following statements generate
proc kde data=outi; ods exclude all; univar beta1 beta2 beta3 beta4 beta5 beta6 / out=m1 (drop=count); run; ods exclude none;
%reshape(m1, mcmc, suffix1=, suffix2=md);
Example 52.8: Cox Models
F
4287
data phout; set phout(drop = LogPost Iteration); beta1 = LogBUN; beta2 = z2; beta3 = HGB; beta4 = z3; beta5 = Platelet; beta6 = z4; drop LogBUN HGB Platelet z2-z4; run; proc kde data=phout; ods exclude all; univar beta1 beta2 beta3 beta4 beta5 beta6 / out=m2 (drop=count); run; ods exclude none;
%reshape(m2, phreg, suffix1=p, suffix2=pd); data all; merge mcmc phreg; run; proc template; define statgraph twobythree;
%macro plot; begingraph; layout lattice / rows=2 columns=3;
%do i = 1 %to 6; layout overlay /yaxisopts=(label=" "); seriesplot y=beta&i.md x=beta&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "MCMC" name="MCMC"; seriesplot y=beta&i.pd x=beta&i.p
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "PHREG" name="PHREG"; endlayout;
%end;
Sidebar / align = bottom; discretelegend "MCMC" "PHREG"; endsidebar; endlayout; endgraph;
%mend; %plot; end; run; proc sgrender data=all template=twobythree; title "Kernel Density Comparison"; run;
The macro %RESHAPE is defined in the example “ Logistic Regression Random-Effects Model ” on
page 4238.
4288
F
Chapter 52: The MCMC Procedure
Output 52.8.6
Comparing Estimates from PROC MCMC and PROC PHREG
You can use PROC MCMC to fit failure time data that can be right, left, or interval censored. To illustrate, a normal regression model is used in this example.
Assume that you have the following simple regression model with no covariates: y
D C where y is a vector of response values (the failure times), is the grand mean, is an unknown scale parameter, and are errors from the standard normal distribution. Instead of observing you only observe a truncated value t i
. If the true y i y i directly, occurs after the censored time t i
, it is called right censoring
. If y i occurs before the censored time, it is called left censoring
. A failure time y i censored at both ends, and this is called interval censoring . The likelihood for y i can be is as follows: p.y
i j / D
ˆ
ˆ
<
ˆ
ˆ
:
8
.y
i j
; /
S.t
1 l;i j /
S.t
r;i j /
S.t
l;i j / S.t
r;i j / if if if if y i y i y i y i is uncensored is right censored by t l;i is left censored by t r;i is interval censored by t l;i and t r;i where
S.
/ is the survival function,
S.t / D P r.T > t /
.
patients treated with radiotherapy:
Example 52.9: Normal Regression with Interval Censoring
F
4289
title 'Normal Regression with Interval Censoring'; data cosmetic; label tl = 'Time to Event (Months)'; input tl tr @@;
45 datalines;
.
37 44
6 10
.
8
25 37
37 .
46
40
.
.
.
32
;
5
.
18
33
.
.
.
7
4 11
26 40
17 25
24 .
19 26
46
15
46
46
36
37
.
.
.
.
.
.
46 .
11 15
27 34
11 18
5 11
34 .
7 16
22
36 44
38
19 35
36
.
.
.
17
46
46
.
.
.
5 12
17 25
7 14
46 .
36 48
37 .
24 .
The data consist of time interval endpoints (in months). Nonmissing equal endpoints ( tl
= tr
) indicates uncensoring; a nonmissing lower endpoint ( tl
¤
.) and a missing upper endpoint ( tr
= .) indicates right censoring; a missing lower endpoint ( tl
= .) and a nonmissing upper endpoint ( left censoring; and nonmissing unequal endpoints ( tl
¤ tr
¤ tr
) indicates interval censoring.
.) indicates
With this data set, you can consider using proper but diffuse priors on both and , for example:
./
. /
/
/
.0; sd
D 1000/ f
.0:001; iscale D 0:001/ where f
is the gamma density function.
The following SAS statements fit an interval censoring model and generate
proc mcmc data=cosmetic outpost=postout seed=1 nmc=20000 missing=AC; ods select PostSummaries PostIntervals; parms mu 60 sigma 50; prior mu ~ normal(0, sd=1000); prior sigma ~ gamma(shape=0.001,iscale=0.001); if (tl^=. and tr^=. and tl=tr) then llike = logpdf('normal',tr,mu,sigma); else if (tl^=. and tr=.) then llike = logsdf('normal',tl,mu,sigma); else if (tl=. and tr^=.) then llike = logcdf('normal',tr,mu,sigma); else llike = log(sdf('normal',tl,mu,sigma) sdf('normal',tr,mu,sigma)); model general(llike); run;
Because there are missing cells in the input data, you want to use the
option so that PROC MCMC does not delete any observations that contain missing values. The IF-ELSE statements distinguish different censoring cases for y i
, according to the likelihood. The SAS functions LOGCDF, LOGSDF, LOGPDF, and SDF are useful here. The
statement assigns
4290
F
Chapter 52: The MCMC Procedure llike as the log likelihood to the response. The Markov chain appears to have converged in this example (evidence not shown here), and the posterior estimates are shown in
Output 52.9.1
Interval Censoring
Parameter mu sigma
Normal Regression with Interval Censoring
The MCMC Procedure
N
20000
20000
Posterior Summaries
Mean
Standard
Deviation
41.7807
29.1122
5.7882
6.0503
25%
Percentiles
50%
37.7220
24.8774
41.3468
28.2210
75%
45.2249
32.4250
Parameter mu sigma
Alpha
0.050
0.050
Posterior Intervals
Equal-Tail Interval
32.0499
20.0889
54.6104
43.1335
HPD Interval
31.3604
19.4041
53.6115
41.6742
Conjoint analysis uses regression techniques to model consumer preferences and to estimate consumer utility functions. A problem with conventional conjoint analysis is that sometimes your estimated utilities do not make sense. Your results might suggest, for example, that the consumers would prefer to spend more on a product than to spend less. With PROC MCMC, you can specify constraints on the part-worth utilities (parameter estimates). Suppose that the consumer product being analyzed is an off-road motorcycle. The relevant attributes are how large each motorcycle is (less than 300cc,
301–550cc, and more than 551cc), how much it costs (less than $5000, $5001–$6000, $6001–$7000, and more than $7000), whether or not it has an electric starter, whether or not the engine is counterbalanced, and whether the bike is from Japan or Europe. The preference variable is a ranking of the bikes. You could perform an ordinary conjoint analysis with PROC TRANSREG (see Chapter 91,
“ The TRANSREG Procedure ”) as follows:
options validvarname=any; proc format; value sizef 1 = '< 300cc' 2 = '300-550cc' 3 = '> 551cc'; value pricef 1 = '< $5000' 2 = '$5000 - $6000'
3 = '$6001 - $7000' 4 = '> $7000'; value startf 1 = 'Electric Start' 2 = 'Kick Start'; value balf value orif run;
1 = 'Counter Balanced' 2 = 'Unbalanced';
1 = 'Japanese' 2 = 'European'; data bikes; input Size Price Start Balance Origin Rank @@; format size sizef. price pricef. start startf.
Example 52.10: Constrained Analysis
F
4291
datalines;
2 1 2 1 2 3
3 3 1 1 2
2 3 2 2 1
2 4 1 1 1
; balance balf. origin orif.;
1
9
4
1 4 2 2 2
1 3 2 1 1
1 1 1 2 1
7
5
8
3 1 1 2 1 11
1 2 1 1 2 6
3 4 2 2 2 12
2 2 1 2 2 10
3 2 2 1 1 2 title 'Ordinary Conjoint Analysis by PROC TRANSREG'; proc transreg data=bikes utilities cprefix=0 lprefix=0; ods select Utilities; model identity(rank / reflect) = class(size price start balance origin / zero=sum); output out=coded(drop=intercept) replace; run;
The DATA step reads the experimental design and dependent variable
Rank and assigns formats to label the factor levels. PROC TRANSREG is run specifying UTILITIES, which requests a conjoint analysis. The rank variable is reflected around its mean (1
!
12, 2
!
11, . . . , 12
!
1) so that in the analysis, larger part-worth utilities correspond to higher preference. The OUT=CODED data set contains the reflected ranks and a binary coding of the factors that can be used in other analyses. Refer to
PROC TRANSREG.
( 2004 ) for more information about conjoint analysis and coding with
The Utilities table from the conjoint analysis is shown in
utilities for price. The part-worth utility for < $5000 is 0.25. As price increases to the $5000–$6000 range, utility decreases to
0:5
. Then as price increases to the $6001–$7000 range, part-worth utility increases to 0.5. Finally, for the most expensive bikes, utility decreases again to 0:25 . In cases like this, you might want to impose constraints on the solution so that the part-worth utility for price never increases as prices go up.
4292
F
Chapter 52: The MCMC Procedure
Output 52.10.1
Ordinary Conjoint Analysis by PROC TRANSREG
Label
Intercept
< 300cc
300-550cc
> 551cc
< $5000
$5000 - $6000
$6001 - $7000
> $7000
Electric Start
Kick Start
Counter Balanced
Unbalanced
Japanese
European
Ordinary Conjoint Analysis by PROC TRANSREG
The TRANSREG Procedure
Utilities Table Based on the Usual Degrees of Freedom
Utility
Standard
Error
Importance
(% Utility
Range) Variable
6.5000
-0.0000
-0.0000
0.0000
0.95743
1.35401
1.35401
1.35401
0.000
Intercept
Class.< 300cc
Class.300-550cc
Class.> 551cc
13.333
0.2500
-0.5000
0.5000
-0.2500
-0.1250
0.1250
3.0000
-3.0000
-0.1250
0.1250
1.75891
1.75891
1.75891
1.75891
1.01550
1.01550
1.01550
1.01550
1.01550
1.01550
3.333
80.000
3.333
Class.< $5000
Class.$5000 - $6000
Class.$6001 - $7000
Class.> $7000
Class.Electric Start
Class.Kick Start
Class.Counter Balanced
Class.Unbalanced
Class.Japanese
Class.European
You could run PROC TRANSREG again, specifying monotonicity constraints on the part-worth utilities for price:
title 'Constrained Conjoint Analysis by PROC TRANSREG'; proc transreg data=bikes utilities cprefix=0 lprefix=0; ods select ConservUtilities; model identity(rank / reflect) = monotone(price / tstandard=center) class(size start balance origin / zero=sum); run;
The output from this PROC TRANSREG step is shown in
Example 52.10: Constrained Analysis
F
4293
Output 52.10.2
Constrained Conjoint Analysis by PROC TRANSREG
Constrained Conjoint Analysis by PROC TRANSREG
The TRANSREG Procedure
Label
Intercept
Price
< $5000
$5000 - $6000
$6001 - $7000
> $7000
Utilities Table Based on Conservative Degrees of Freedom
Utility
Standard
Error
Importance
(% Utility
Range) Variable
6.5000
-0.1581
0.2500
0.0000
0.0000
-0.2500
0.97658
.
.
.
.
.
7.143
Intercept
Monotone(Price)
0.000
< 300cc
300-550cc
> 551cc
Electric Start
Kick Start
Counter Balanced
Unbalanced
Japanese
European
-0.0000
0.0000
0.0000
-0.2083
0.2083
3.0000
-3.0000
-0.0417
0.0417
1.38109
1.38109
1.38109
1.00663
1.00663
0.97658
0.97658
1.00663
1.00663
5.952
85.714
1.190
Class.< 300cc
Class.300-550cc
Class.> 551cc
Class.Electric Start
Class.Kick Start
Class.Counter Balanced
Class.Unbalanced
Class.Japanese
Class.European
This monotonicity constraint is one of the few constraints on the part-worth utilities that you can specify in PROC TRANSREG. In contrast, PROC MCMC allows you to specify any constraint that can be written in the DATA step language. You can perform the restricted conjoint analysis with
PROC MCMC by using the coded factors that were output from PROC TRANSREG. The data set is coded
.
The likelihood is a simple regression model: rank i normal
.
x
0 i
ˇ
; / where rank is the response, the covariates are
‘< 300cc’n
,
‘300-500cc’n
,
‘< $5000’n
,
‘$5000 - $6000’n
,
‘$6001 - $7000’n
,
‘Electric Start’n
,
‘Counter Balanced’n
, and
Japanese
. Note that OPTIONS VALIDVAR-
NAME=ANY allows PROC TRANSREG to create names for the coded variables with blanks and special characters. That is why the name-literal notation (‘ variable-name
’n) is used for the input data set variables.
Suppose that there are two constraints you want to put on some of the parameters: one is that the parameters for
‘< $5000’n
,
‘$5000 - $6000’n
, and
‘$6001 - $7000’n decrease in order, and the other is that the parameter for
‘Counter Balanced’n is strictly positive. You can consider a truncated multivariate normal prior as follows:
ˇ
‘< $5000’n
; ˇ
‘$5000 - $6000’n
; ˇ
‘$6001 - $7000’n
; ˇ
‘Counter Balanced’n
MVN
.0; I /
4294
F
Chapter 52: The MCMC Procedure with the following set of constraints:
ˇ
‘< $5000’n
ˇ
‘Counter Balanced’n
> ˇ
‘$5000 - $6000’n
> 0
> ˇ
‘$6001 - $7000’n
> 0
The condition that
ˇ
‘$6001 - $7000’n utility for the highest price range,
> 0
> $7000 reflects an implied constraint that, by definition,
0 is the
, which is the reference level for the binary coded price variable. The following statements fit the desired model:
title 'Bayesian Constrained Conjoint Analysis by PROC MCMC'; proc mcmc data=coded outpost=bikesout ntu=3000 nmc=50000 thin=10 seed=448; ods select PostSummaries; array sigma[4,4] sigma1-sigma16; array mu[4] mu1-mu4; begincnst; call identity(sigma); call mult(sigma, 100, sigma); call zeromatrix(mu); rc = logmpdfsetsq('v', of sigma1-sigma16); endcnst; parms intercept pw300cc pw300_550cc pwElectricStart pwJapanese ltau 1; parms pw5000 0.3 pw5000_6000 0.2 pw6001_7000 0.1 pwCounterBalanced 1; beginnodata; prior intercept pw300: pwE: pwJ: ~ normal(0, var=100); if (pw5000 >= pw5000_6000 & pw5000_6000 >= pw6001_7000 & pw6001_7000 >= 0 & pwCounterBalanced > 0) then lp = logmpdfnormal(of mu1-mu4, pw5000, pw5000_6000, pw6001_7000, pwCounterBalanced, 'v'); else lp = .; prior pw5000 pw5000_6000 pw6001_7000 pwC: ~ general(lp); prior ltau ~ egamma(0.001, scale=1000); tau = exp(ltau); endnodata; mean = intercept + pw300cc
* '< 300cc'n
+ pw300_550cc
* '300-550cc'n
+ pw5000
* '< $5000'n
+ pw5000_6000
* '$5000 - $6000'n
+ pw6001_7000
* '$6001 - $7000'n
+ pwElectricStart
* 'Electric Start'n
+ pwCounterBalanced * 'Counter Balanced'n + pwJapanese
* Japanese; model rank ~ normal(mean, prec=tau); run; data _null_;
Example 52.10: Constrained Analysis
F
4295
rc = logmpdffree(); run;
The two
statements allocate a
4 array of size 4 for the prior means. In the
IDENTITY function sets sigma
4 dimensional array for the prior covariance and an
and
statements, the CALL to be an identity matrix; the CALL MULT function sets sigma
’s diagonal elements to be 100 (the diagonal variance terms); the CALL ZEROMATRIX function sets mu to be a vector of zeros (the prior means); and the LOGMPDFSETSQ function sets up sigma to be called in a multivariate normal density function later. For matrix functions in PROC MCMC, see the
section “ Matrix Functions in PROC MCMC ” on page 4176. For multivariate density functions, see
the LOGMPDFSET or the LOGMPDFSETSQ functions to set up covariance matrix, you must free the memory allocated by these functions after you exit PROC MCMC. To free the memory, use the function LOGMPDFFREE.
There are two
statements, with each of them naming a block of parameters. The first
statement blocks the following: the intercept, the two size parameters, the one start-type parameter, the one origin parameter, and the log of the precision. The second
statement blocks the three price parameters and the one balance parameter, parameters that have the constraint multivariate normal prior. The second
statement also specifies initial values for the parameter estimates.
The initial values reflect the constraints on these parameters. The initial part-worth utilities all decrease from 0.3 to 0.2 to 0.1 to 0.0 (for the implicit reference level) as the prices increase. Also, the initial part-worth utility for the counter-balanced engine is set to a positive value, 1.
In the
statements, regression coefficients without constraints are given an independent normal prior with mean at 0 and variance of 100. The next IF-ELSE construction imposes the constraints.
When these constraints are met, pw5000, pw5000_6000, pw6001_7000, pwCounterBalanced are jointly distributed as a multivariate normal prior with mean mu and covariance sigma
(as defined via the symbol
‘v’ in the
and
lp is assigned a missing value.
statements). Otherwise, the prior is not defined and
The parameter ltau is given an egamma prior. It is an equivalent prior to placing a gamma prior, with the same configuration, on tau
D exp
.
ltau
/
. For the definition of the egamma distribution, see the
section “ Standard Distributions ” on page 4155. This transformation often improves mixing (see
“ Example 52.4: Nonlinear Poisson Regression Models
” on page 4229 and “ Example 52.12: Using a
Transformation to Improve Mixing ” on page 4307). The next assignment statement transforms
ltau back to tau
.
The model specification is linear. The mean is comprised of an intercept and the sum of terms like pw300cc * ‘< 300cc’n
, which is a parameter times an input data set variable. The
statement specifies that the linear model for rank is normally distributed with mean mean and precision tau
.
After the PROC MCMC run, you must run the memory clean up function LOGMPDFFREE, which should produce the following note in the log file:
NOTE: The matrix v - has been deleted.
The MCMC results are shown in
4296
F
Chapter 52: The MCMC Procedure
Output 52.10.3
MCMC Results
Bayesian Constrained Conjoint Analysis by PROC MCMC
The MCMC Procedure
Parameter intercept pw300cc pw300_550cc pwElectricStart pwJapanese ltau pw5000 pw5000_6000 pw6001_7000 pwCounterBalanced
N
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
Posterior Summaries
Mean
Standard
Deviation
2.2052
0.0780
-0.0173
-1.2175
-0.4212
-2.4440
4.3724
2.6649
1.4880
5.9056
2.6285
2.5670
2.5378
2.1805
2.1485
0.7293
2.4962
1.8227
1.3303
2.0591
25%
Percentiles
50% 75%
0.8089
-1.4062
-1.5136
-2.4933
-1.6575
-2.9024
2.6418
1.3878
0.5077
4.6440
2.3658
0.0717
-0.00275
-1.1041
-0.4102
-2.3787
3.9163
2.2894
1.1389
5.9033
3.8732
1.5850
1.4536
0.1410
0.7909
-1.9177
5.5202
3.5162
2.0849
7.1036
The estimates of the part-worth utility for the price categories are ordered as expected. This agrees with the intuition that there is a higher preference for a less expensive motor bike when all other things are equal, and that is what you see when you look at the estimated posterior means for the price part-worths. The estimated standard deviations of the price part-worths in this model are of approximately the same order of magnitude as the posterior means. This indicates that the part-worth utilities for this subject are not significantly far from each other, and that this subject’s ranking of the options was not significantly influenced by the difference in price.
One advantage of Bayesian analysis is that you can incorporate prior information in the data analysis.
Constraints on the parameter space are one possible source of information that you might have before you examine the data. This example shows that it can easily be accomplished in PROC MCMC.
This example illustrates using the
statement to implement a new Markov chain sampler. The algorithm demonstrated here is proposed by
Holmes and Held ( 2006 ), hereafter referred to as HH.
They presented a Gibbs sampling algorithm for generating draws from the posterior distribution of the parameters in a probit regression model. The notation follows closely to HH.
The data used here is the remission data set from a PROC LOGISTIC example:
title 'Implement a New Sampling Algorithm'; data inputdata; input remiss cell smear infil li blast temp; ind = _n_; cnst = 1; label remiss='Complete Remission';
Example 52.11: Implement a New Sampling Algorithm
F
4297
datalines;
... more lines ...
0 1 0.73
0.73
0.7
0.398
0.986
;
The variable remiss is the cancer remission indicator variable with a value of 1 for remission and a value of 0 for nonremission. There are six explanatory variables: cell
, smear
, infil
, li
, blast
, and temp
These variables are the risk factors thought to be related to cancer remission. The binary regression
.
model is as follows: remiss i binary .p
i
/ where the covariates are linked to p i through a probit transformation: probit
.p
i
/ D x
0
ˇ
ˇ are the regression coefficients and x
0 the explanatory variables. Suppose that you want to use independent normal priors on the regression coefficients:
ˇ i normal
.0; var
D 25/
Fitting a logistic model with PROC MCMC is straightforward. You can use the following statements:
proc mcmc data=inputdata nmc=100000 propcov=quanew seed=17 outpost=mcmcout; ods select PostSummaries ess; parms beta0-beta6; prior beta: ~ normal(0,var=25); mu = beta0 + beta1*cell + beta2*smear + beta3*infil + beta4*li + beta5*blast + p = cdf('normal', mu, 0, 1); model remiss ~ bern(p); run; beta6*temp;
The expression mu is the regression mean, and the CDF function links remission p in the binary likelihood.
mu to the probability of
The summary statistics and effective sample sizes tables are shown in
autocorrelations among the posterior samples, and efficiency is relatively low. The correlation time is reduced only after a large amount of thinning.
4298
F
Chapter 52: The MCMC Procedure
Output 52.11.1
Random Walk Metropolis
Parameter beta0 beta1 beta2 beta3 beta4 beta5 beta6
N
100000
100000
100000
100000
100000
100000
100000
Implement a New Sampling Algorithm
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-2.0531
2.6300
-0.8426
1.5933
2.0390
-0.3184
-3.2611
3.8299
2.8270
3.2108
3.5491
0.8796
0.9543
3.7806
25%
Percentiles
50%
-4.6418
0.6563
-3.0270
-0.7993
1.4312
-0.9613
-5.8050
-2.0354
2.5272
-0.8263
1.6190
2.0028
-0.3123
-3.2736
75%
0.5638
4.4846
1.3429
3.9695
2.6194
0.3418
-0.7243
Parameter beta0 beta1 beta2 beta3 beta4 beta5 beta6
Implement a New Sampling Algorithm
The MCMC Procedure
Effective Sample Sizes
ESS
Autocorrelation
Time
4280.8
4496.5
3434.1
3856.6
3659.7
3229.9
4430.7
23.3602
22.2398
29.1199
25.9294
27.3245
30.9610
22.5696
Efficiency
0.0428
0.0450
0.0343
0.0386
0.0366
0.0323
0.0443
As an alternative to the random walk Metropolis, you can use the Gibbs algorithm to sample from
the posterior distribution. The Gibbs algorithm is described in the section “ Gibbs Sampler ” on
page 151. While the Gibbs algorithm generally applies to a wide range of statistical models, the actual implementation can be problem-specific. In this example, performing a Gibbs sampler involves introducing a class of auxiliary variables (also known as latent variables). You first reformulate the model by adding a z i for each observation in the data set: y z
ˇ i i
D
D
1
0 if z i
> 0 otherwise x
0 i
ˇ
C i normal
.0; 1/
.
ˇ
/
If ˇ has a normal prior, such as full conditional distribution of ˇ
.
ˇ
/ D N.
b
; v
/
, you can work out a closed form solution to the given the data and the latent variables z i
. The full conditional
Example 52.11: Implement a New Sampling Algorithm
F
4299 distribution is also a multivariate normal, due to the conjugacy of the problem. See the section
“ Conjugate Priors ” on page 144. The formula is shown here:
ˇ j z
; x
B
V
D
D normal
.
B
;
V
/
V ..v/
.
v
1
1 b
C x
0 x
C x
0 z /
/
1
The advantage of creating the latent variables is that the full conditional distribution of z is also easy to work with. The distribution is a truncated normal distribution: z i j
ˇ ; x i
; y i normal
.
x i
ˇ
; 1/I.z
i normal .
x i
ˇ ; 1/I.z
i
> 0/
0/ if y i
D 1 otherwise
You can sample ˇ and z iteratively, by drawing high degree of correlation could exist between ˇ
ˇ and given z and vice verse. HH point out that a z
, and it makes this iterative way of sampling inefficient. As an improvement, HH proposed an algorithm that samples ˇ and z jointly. At each iteration, you sample z i from the posterior marginal distribution (this is the distribution that is conditional only on the data and not on any parameters) and then sample ˇ from the same posterior full conditional distribution as described previously:
1. Sample z i from its posterior marginal distribution: z i j z i
; y i m i v i w i h i
D
D
D
D normal
.m
i
; v i
/I.z
i normal .m
i
; v i
/I.z
i
> 0/
0/ x i
B w i
.z
i x i
B /
1 C w i h i
=.1
h i
/
.
H / i i; H D xVx
0 if y i
D 1 otherwise
2. Sample ˇ from the same posterior full conditional distribution described previously.
For a detailed description of each of the conditional terms, refer to the original paper.
PROC MCMC cannot sample from the probit model by using this sampling scheme but you can implement the algorithm by using the
statement. To sample z i from its marginal, you need a function that draws random variables from a truncated normal distribution. The functions,
RLTNORM and RRTNORM, generate left- and right-truncated normal variates, respectively. The algorithm is taken from
The functions are written in PROC FCMP (see the FCMP Procedure in the
Base SAS Procedures
Guide ):
proc fcmp outlib=sasuser.funcs.uds;
/******************************************/
/* Generate left-truncated normal variate */
/******************************************/ function rltnorm(mu,sig,lwr);
4300
F
Chapter 52: The MCMC Procedure
if lwr<mu then do; ans = lwr-1; do while(ans<lwr); ans = rand('normal',mu,sig); end; end; else do; mul = (lwr-mu)/sig; alpha = (mul + sqrt(mul**2 + 4))/2; accept=0; do while(accept=0); z = mul + rand('exponential')/alpha; lrho = -(z-alpha)**2/2; u = rand('uniform'); lu = log(u); if lu <= lrho then accept=1; end; ans = sig*z + mu; end; return(ans); endsub;
/*******************************************/
/* Generate right-truncated normal variate */
/*******************************************/ function rrtnorm(mu,sig,uppr); ans = 2*mu - rltnorm(mu,sig, 2*mu-uppr); return(ans); endsub; run;
The function call to RLTNORM(mu,sig,lwr) generates a random number from the left-truncated normal distribution: normal
.
mu
; sd
D sig
/I. > lwr
/
Similarly, the function call to RRTNORM(mu,sig,uppr) generates a random number from the righttruncated normal distribution: normal
.
mu
; sd
D sig
/I. < lwr
/
These functions are used to generate the latent variables z i
.
Using the algorithm A1 from the HH paper as an example,
lists a line-by-line implementation with the PROC MCMC coding style. The table is broken into three portions: set up the constants, initialize the parameters, and sample one draw from the posterior distribution. The left column of the table is identical to the A1 algorithm stated in the appendix of HH. The right column of the table lists SAS statements.
Example 52.11: Implement a New Sampling Algorithm
F
4301
Table 52.37
Holmes and Held (2006), algorithm A1. Side-by-Side Comparison to SAS
Define Constants In the BEGINCNST/ENDCNST Statements
V .X
T
X C v
1
/
1
call transpose(x,xt); /* xt = transpose(x) */ call mult(xt,x,xtx); call inv(v,v); /* v = inverse(v) */ call addmatrix(xtx,v,xtx); /* xtx = xtx+v */ call inv(xtx,v); /* v = inverse(xtx) */
L
Chol
.V /
S V X
T
call chol(v,L); call mult(v,xt,S);
FOR j
H Œj
D 1 to n
X Œj; S Œ; j
W Œj
QŒj
H Œj =.1
W Œj C 1
END
H Œj /
call mult(x,S,HatMat); do j=1 to &n;
H = HatMat[j,j];
W[j] = H/(1-H); sQ[j] = sqrt(W[j] + 1); /* use s.d.
end; in SAS */
Initial Values
Z
B normal
SZ
.0; I n
/I nd.Y; Z/
In the BEGINCNST/ENDCNST Statements
do j=1 to &n; if(y[j]=1) then
Z[j] = rltnorm(0,1,0); else
Z[j] = rrtnorm(0,1,0); end; call mult(S,Z,B);
4302
F
Chapter 52: The MCMC Procedure
Table 52.37
(continued)
Draw One Sample Subroutine HH
do j=1 to &n; zold = Z[j]; m = 0; do k= 1 to &p; m = m + X[j,k] * B[k]; end;
FOR z m j old
D 1 to n
ZŒj
X Œj; B m
ZŒj m W Œj .ZŒj normal m/
.m; QŒj /I nd.Y Œj ; ZŒj /
B
END
B C .ZŒj z old
/S Œ; j
T
ˇŒ; i normal
.0; I p
/
B C LT
m = m - W[j]*(Z[j]-m); if (y[j]=1) then
Z[j] = rltnorm(m,sQ[j],0); else
Z[j] = rrtnorm(m,sQ[j],0); diff = Z[j] - zold; do k= 1 to &p;
B[k] = B[k] + diff * S[k,j]; end; end; do j = 1 to &p;
T[j] = rand(’normal’); end; call mult(L,T,T); call addmatrix(B,T,beta);
The following statements define the subroutine HH (algorithm A1) in PROC FCMP and store it in library
sasuser.funcs.uds
:
/* define the HH algorithm in PROC FCMP. */
%let n = 27;
%let p = 7; options cmplib=sasuser.funcs; proc fcmp outlib=sasuser.funcs.uds; subroutine HH(beta[*],Z[*],B[*],x[*,*],y[*],W[*],sQ[*],S[*,*],L[*,*]); outargs beta, Z, B; array T[&p] / nosym; do j=1 to &n; zold = Z[j]; m = 0; do k = 1 to &p; m = m + X[j,k] * B[k]; end; m = m - W[j]*(Z[j]-m); if (y[j]=1) then
Z[j] = rltnorm(m,sQ[j],0); else
Z[j] = rrtnorm(m,sQ[j],0); diff = Z[j] - zold; do k = 1 to &p;
B[k] = B[k] + diff * S[k,j];
Example 52.11: Implement a New Sampling Algorithm
F
4303
end; end; do j=1 to &p;
T[j] = rand('normal'); end; call mult(L,T,T); call addmatrix(B,T,beta); endsub; run;
Note that one-dimensional array arguments take the form of name [*] and two-dimensional array arguments take the form of name [*,*]. Three variables, beta
,
Z
, and
B
, are OUTARGS variables, making them the only arguments that can be modified in the subroutine. For the
to work, all OUTARGS variables have to be model parameters. Technically, only beta statement and
Z are model parameters, and
B is not. The reason that
B is declared as an OUTARGS is because the array must be updated throughout the simulation, and this is the only way to modify its values. The input array x contains all of the explanatory variables, and the array y stores the response. The rest of the input arrays,
W
, sQ
,
S
, and
L
, store constants as detailed in the algorithm. The following statements illustrate how to fit a Bayesian probit model by using the HH algorithm:
options cmplib=sasuser.funcs; proc mcmc data=inputdata nmc=5000 monitor=(beta) outpost=hhout; ods select PostSummaries ess; array xtx[&p,&p]; array xt[&p,&n]; array v[&p,&p]; array HatMat[&n,&n];
/* work space
/* work space
/* work space
/* work space
/* V * Xt
*/
*/
*/
*/ array S[&p,&n]; array W[&n];
*/ array y[1]/ nosymbols; /* y stores the response variable
*/ array x[1]/ nosymbols; /* x stores the explanatory variables */ array sQ[&n]; array B[&p];
/* sqrt of the diagonal elements of Q */
/* conditional mean of beta
*/ array L[&p,&p]; array Z[&n];
/* Cholesky decomp of conditional cov */
/* latent variables Z
*/ array beta[&p] beta0-beta6; /* regression coefficients
*/ begincnst; call streaminit(83101); if ind=1 then do; rc = read_array("inputdata", x, "cnst", "cell", "smear", "infil",
"li", "blast", "temp"); rc = read_array("inputdata", y, "remiss"); call identity(v); call mult(v, 25, v); call transpose(x,xt); call mult(xt,x,xtx); call inv(v,v); call addmatrix(xtx,v,xtx); call inv(xtx,v); call chol(v,L); call mult(v,xt,S);
4304
F
Chapter 52: The MCMC Procedure
call mult(x,S,HatMat); do j=1 to &n;
H = HatMat[j,j];
W[j] = H/(1-H); sQ[j] = sqrt(W[j] + 1); end; do j=1 to &n; if(y[j]=1) then else
Z[j] = rltnorm(0,1,0);
Z[j] = rrtnorm(0,1,0); end; call mult(S,Z,B); end; endcnst; uds HH(beta,Z,B,x,y,W,sQ,S,L); parms z: beta: 0 B1-B7 / uds; prior z: beta: B1-B7 ~ general(0); model general(0); run;
The OPTIONS statement names the catalog of FCMP subroutines to use. The cmplib library stores the subroutine HH. You do not need to set a random number seed in the PROC MCMC statement because all random numbers are generated from the HH subroutine. The initialization of the rand function is controlled by the streaminit function, which is called in the program with a seed value of
83101.
A number of arrays are allocated. Some of them, such as xtx
, xt
, v
, and
HatMat
, allocate work space that is used to construct constant arrays. Other arrays are used in the subroutine sampling.
Explanations of the arrays are shown in comments in the statements.
In the
and
statement block, you read data set variables in the arrays x and y
, calculate all the constant terms, and assign initial values to
Z and
B
. For the READ_ARRAY
function, see the section “ READ_ARRAY Function ” on page 4134. For listings of all array functions
and their definitions, see the section “ Matrix Functions in PROC MCMC ” on page 4176.
The
statement declares that the subroutine HH is used to sample the parameters
B
. You also specify the
option in the
beta
,
Z
, and statement. Because all parameters are updated through the
interface, it is not necessary to declare the actual form of the prior for any of the parameters. Each parameter is declared to have a prior of general(0)
. Similarly, it is not necessary to declare the actual form of the likelihood. The
statement also takes a flat likelihood of the form
general(0)
.
Summary statistics and effective sample sizes are shown in
are very close to what was shown in
. The HH algorithm produces samples that are
much less correlated.
Example 52.11: Implement a New Sampling Algorithm
F
4305
Output 52.11.2
Holms and Held
Parameter beta0 beta1 beta2 beta3 beta4 beta5 beta6
N
5000
5000
5000
5000
5000
5000
5000
Implement a New Sampling Algorithm
The MCMC Procedure
Posterior Summaries
Mean
Standard
Deviation
-2.0567
2.7254
-0.8318
1.6319
2.0567
-0.3473
-3.3787
3.8260
2.8079
3.2017
3.5108
0.8800
0.9490
3.7991
25%
Percentiles
50%
-4.6537
0.7812
-2.9987
-0.7481
1.4400
-0.9737
-5.9089
-2.0777
2.6678
-0.8626
1.6636
2.0266
-0.3267
-3.3504
75%
0.5495
4.5370
1.2918
4.0302
2.6229
0.2752
-0.7928
Parameter beta0 beta1 beta2 beta3 beta4 beta5 beta6
Implement a New Sampling Algorithm
The MCMC Procedure
Effective Sample Sizes
ESS
Autocorrelation
Time
3651.3
1563.8
5005.9
4853.2
2611.2
3049.2
3503.2
1.3694
3.1973
0.9988
1.0302
1.9148
1.6398
1.4273
Efficiency
0.7303
0.3128
1.0012
0.9706
0.5222
0.6098
0.7006
The following statements generate the kernel density comparison plots shown in
proc kde data=mcmcout; ods exclude all; univar beta0 beta1 beta2 beta3 beta4 beta5 beta6 / out=m1(drop=count); run; ods exclude none;
%reshape(m1, mcmc, suffix1=, suffix2=md); proc kde data=hhout(drop = LogPost logprior loglike Iteration z1-z27 b1-b7);; ods exclude all; univar beta0 beta1 beta2 beta3 beta4 beta5 beta6
/ out=m2 (drop=count); run; ods exclude none;
%reshape(m2, hh, suffix1=p, suffix2=pd);
4306
F
Chapter 52: The MCMC Procedure
data all; merge mcmc hh; run; proc template; define statgraph threebythree;
%macro plot; begingraph; layout lattice / rows=3 columns=3;
%do i = 0 %to 6; layout overlay /yaxisopts=(label=" "); seriesplot y=beta&i.md x=beta&i
/ connectorder=xaxis lineattrs=(pattern=mediumdash color=blue) legendlabel = "PROC MCMC" name="MCMC"; seriesplot y=beta&i.pd x=beta&i.p
/ connectorder=xaxis lineattrs=(color=red) legendlabel = "Holmes and Held" name="HH"; endlayout;
%end;
Sidebar / align = bottom; discretelegend "MCMC" "HH"; end; run; endsidebar; endlayout; endgraph;
%mend; %plot; proc sgrender data=all template=threebythree; title "Kernel Density Comparison"; run;
The macro %RESHAPE is defined in the example “ Logistic Regression Random-Effects Model ” on
page 4238.
Example 52.12: Using a Transformation to Improve Mixing
F
4307
Output 52.11.3
Kernel Density Comparison
It is interesting to compare the two approaches of fitting a generalized linear model. The random walk
Metropolis on a seven-dimensional parameter space produces autocorrelations that are substantially higher than the HH algorithm. A much longer chain is needed to produce roughly equivalent effective sample sizes. On the other hand, the Metropolis algorithm is faster to run. The running time of these two examples is roughly the same, with the random walk Metropolis with 100000 samples, a 20-fold increase over that in the HH algorithm example. The speed difference can be attributed to a number of factors, ranging from the implementation of the software and the overhead cost of calling PROC
FCMP subroutine and functions. In addition, the HH algorithm requires more parameters by creating an equal number of latent variables as the sample size. Sampling more parameters takes time. A larger number of parameters also increases the challenge in convergence diagnostics, because it is imperative to have convergence in all parameters before you make valid posterior inferences. Finally, you might feel that coding in PROC MCMC is easier. However, this really is not a fair comparison to make here. Writing a Metropolis algorithm from scratch would have probably taken just as much, if not more, effort than the HH algorithm.
Proper transformations of parameters can often improve the mixing in PROC MCMC. You already
saw this in “ Example 52.4: Nonlinear Poisson Regression Models ” on page 4229, which sampled
using the log scale of parameters that priors that are strictly positive, such as the gamma priors. This
4308
F
Chapter 52: The MCMC Procedure example shows another useful transformation: the logit transformation on parameters that take a uniform prior on [0, 1].
The data set is taken from
Sharples ( 1990 ). It is used in
Chaloner and Brant ( 1988 ) and Chaloner
( 1994 ) to identify outliers in the data set in a two-level hierarchical model.
this data set to demonstrates the same technique. This example uses the data set to illustrate how mixing can be improved using transformation and does not address the question of outlier detection as in those papers. The following statements create the data set:
data inputdata; input nobs grp y @@; ind = _n_; datalines;
1 1 24.80 2 1 26.90 3 1 26.65
4 1 30.93 5 1 33.77 6 1 63.31
1 2 23.96 2 2 28.92 3 2 28.19
4 2 26.16 5 2 21.34 6 2 29.46
1 3 18.30 2 3 23.67 3 3 14.47
4 3 24.45 5 3 24.89 6 3 28.95
1 4 51.42 2 4 27.97 3 4 24.76
4 4 26.67 5 4 17.58 6 4 24.29
1 5 34.12 2 5 46.87 3 5 58.59
4 5 38.11 5 5 47.59 6 5 44.67
;
There are five groups ( grp
, j D 1; ; 5
) with six observations ( two-level hierarchical model is specified as follows: nobs
, i D 1; ; 6
) in each. The y ij j p normal
.
j
; prec
D normal .; prec
D w b
/
/ normal .0; prec
D
1e 6/ gamma
.0:001; iscale
D 0:001/ uniform
.0; 1/ with the precision parameters related to each other in the following way: b w
D
D
=p b
The total number of parameters in this model is eight:
1
; ;
5
; ;
, and p
.
The following statements fit the model:
ods graphics on; proc mcmc data=inputdata nmc=50000 thin=10 outpost=m1 seed=17 plot=trace; ods select ess tracepanel; array theta[5]; parms theta:;
Example 52.12: Using a Transformation to Improve Mixing
F
4309
parms p tau; parms mu ; beginnodata; hyper p ~ uniform(0,1); hyper tau ~ gamma(shape=0.001,iscale=0.001); hyper mu ~ normal(0,prec=0.00000001); taub = tau/p; prior theta: ~ normal(mu,prec=taub); tauw = taub-tau; endnodata; model y ~ normal(theta[grp],prec=tauw); run;
The ODS SELECT statement displays the effective sample size table and the trace plots. The ODS
GRAPHICS ON statement requests ODS Graphics. The PROC MCMC statement specifies the usual
options for the MCMC run and produces trace plots ( PLOTS=TRACE ). The
statement allocates an array of size 5 for theta
. The three
statements put parameters in three different blocks. The remaining statements specify the hyperprior, prior, and likelihood for the data, as described previously. The resulting trace plots are shown in
sizes table is shown in
Output 52.12.1
Trace Plots
4310
F
Chapter 52: The MCMC Procedure
Output 52.12.1
continued
Example 52.12: Using a Transformation to Improve Mixing
F
4311
Output 52.12.2
Bad Effective Sample Sizes
Parameter theta1 theta2 theta3 theta4 theta5 p tau mu
The MCMC Procedure
Effective Sample Sizes
ESS
Autocorrelation
Time
2207.5
1713.5
1458.5
1904.4
585.9
77.2
140.8
3340.3
2.2650
2.9180
3.4281
2.6255
8.5345
64.7758
35.5052
1.4969
Efficiency
0.4415
0.3427
0.2917
0.3809
0.1172
0.0154
0.0282
0.6681
The trace plots show that most parameters have relatively good mixing. Two exceptions appear to be p and . The trace plot of p shows a slow periodic movement. The parameter does not have good mixing either. When the values are close to zero, the chain stands there for long periods of time. An inspection of the effective sample sizes table reveals the same conclusion: p and have much smaller ESSs than the rest of the parameters.
A scatter plot of the posterior samples of p and reveals why mixing is bad in these two dimensions.
The following statements generate the scatter plot in
title 'Scatter Plot of Parameters on Original Scales'; proc sgplot data=m1; yaxis label = 'p'; xaxis label = 'tau' values=(0 to 0.4 by 0.1); scatter x = tau y = p; run;
4312
F
Chapter 52: The MCMC Procedure
Output 52.12.3
Scatter Plot of versus p
The two parameters clearly have a nonlinear relationship. It is not surprising that the Metropolis algorithm does not work well here. The algorithm is designed for cases where the parameters are linearly related with each other.
Instead of sampling on , you can sample on the log of . The formulation
. /
.
log
. //
/
/ f
.0:001; iscale
D 0:001/ f e
.0:001; iscale
D 0:001/ where f
and f e are density functions for the gamma and egamma distributions. See the section
sample on the logit of p
. The formulation
.p/
/ f uniform
.0; 1/ is equivalent to the following transformation: lgp
.
lgp
/
D
/ logit .p/ exp
.
lgp
/.1
C exp
.
lgp
//
2
Example 52.12: Using a Transformation to Improve Mixing
F
4313
The following statements fit the same model by using transformed parameters:
proc mcmc data=inputdata nmc=50000 thin=10 outpost=m2 seed=17 monitor=(tau p mu theta) plot=trace; ods select ess tracepanel; array theta[5]; parms theta:; parms lgp 0 ltau ; parms mu ; beginnodata; prior ltau ~ egamma(shape=0.001,iscale=0.001); lp = -lgp - 2*log(1+exp(-lgp)); prior lgp ~ general(lp); tau = exp(ltau); p = (1+exp(-lgp))**-1; prior mu ~ normal(0,prec=0.00000001); taub = tau/p; prior theta: ~ normal(mu,prec=taub); tauw = taub-tau; endnodata; model y ~ normal(theta[grp],prec=tauw); run; ods graphics off;
The variable lgp is the logit transformation of p , and ltau is the log transformation of . The prior for ltau is egamma. The lp assignment statement evaluates the log density of
.
lgp
/
. The tau and p assignment statements transform the parameters back to their original scales. The prior distributions for mu
, theta
, and the log likelihood in the
statement remain unchanged. Trace plots
) and effective sample size ( Output 52.12.5
) both show significant improvements in
the mixing for both p and .
4314
F
Chapter 52: The MCMC Procedure
Output 52.12.4
Trace Plots after Transformation
Output 52.12.4
continued
Example 52.12: Using a Transformation to Improve Mixing
F
4315
Output 52.12.5
Effective Sample Sizes after Transformation
Parameter tau p mu theta1 theta2 theta3 theta4 theta5
The MCMC Procedure
Effective Sample Sizes
ESS
Autocorrelation
Time
1916.5
2468.7
3273.9
2184.5
1938.1
1947.1
2115.8
2152.0
2.6089
2.0253
1.5272
2.2888
2.5799
2.5679
2.3632
2.3235
Efficiency
0.3833
0.4937
0.6548
0.4369
0.3876
0.3894
0.4232
0.4304
The following statements generate
and
4316
F
Chapter 52: The MCMC Procedure
title 'Scatter Plot of Parameters on Transformed Scales'; proc sgplot data=m2; yaxis label = 'logit(p)'; xaxis label = 'log(tau)'; scatter x = ltau y = lgp; run; title 'Scatter Plot of Parameters on Original Scales'; proc sgplot data=m2; yaxis label = 'p'; xaxis label = 'tau'; scatter x = tau y = p; run;
Output 52.12.6
Scatter Plot of log . / versus logit
.p/
, After Transformation
Example 52.13: Gelman-Rubin Diagnostics
F
4317
Output 52.12.7
Scatter Plot of versus p
, After Transformation
The scatter plot of log . / versus logit .p/ shows a linear relationship between the two transformed parameters, and this explains the improvement in mixing. In addition, the transformations also help the Markov chain better explore in the original parameter space.
shows a scatter plot of versus p . The plot is similar to
tau has a far longer tail in
, extending all the way to 0.4 as opposed to 0.15 in
that the second Markov chain can explore this dimension of the parameter more efficiently, and as a result, you are able to draw more precise inference with an equal number of simulations.
PROC MCMC does not have the Gelman-Rubin test (see the section “ Gelman and Rubin Diagnostics ”
on page 160) as a part of its diagnostics. The Gelman-Rubin diagnostics rely on parallel chains to test whether they all converge to the same posterior distribution. This example demonstrates how you can
carry out this convergence test. The regression model from the section “ Simple Linear Regression ”
on page 4104 is used. The model has three parameters: and
2 is the variance of the error distribution.
ˇ
0 and
ˇ
1 are the regression coefficients,
The following statements generate the data set:
4318
F
Chapter 52: The MCMC Procedure
title 'Simple Linear Regression, Gelman-Rubin Diagnostics'; data Class; input Name $ Height Weight @@; datalines;
Alfred 69.0 112.5
Carol
Jane
John
Louise
Robert
62.8 102.5
59.8
59.0
56.3
84.5
99.5
77.0
64.8 128.0
William 66.5 112.0
Alice
Henry
Janet
Joyce
Mary
56.5
51.3
84.0
63.5 102.5
62.5 112.5
50.5
66.5 112.0
Ronald 67.0 133.0
;
Barbara 65.3
James 57.3
Jeffrey 62.5
Judy 64.3
Philip
Thomas
98.0
83.0
84.0
90.0
72.0 150.0
57.5
85.0
To run a Gelman-Rubin diagnostic test, you want to start Markov chains at different places in the parameter space. Suppose that you want to start ˇ
0 at 10 , 15 , and 0 ; ˇ
1 at 5 , 10 , and 0 ; and
2 at
1
,
20
, and
50
. You can put these starting values in the following init SAS data set:
data init; input Chain beta0 beta1 sigma2; datalines;
1
2
3
10
-15
0
-5
10
0
1
20
50
;
The following statements run PROC MCMC three times, each with starting values specified in the data set init :
/* define constants */
%let nchain = 3;
%let nparm = 3;
%let nsim = 50000;
%let var = beta0 beta1 sigma2;
%macro gmcmc;
%do i=1 %to &nchain; data _null_; set init; if Chain=&i;
%do j = 1 %to &nparm; call symputx("init&j", %scan(&var, &j));
%end; stop; run; proc mcmc data=class outpost=out&i init=reinit nbi=0 nmc=&nsim stats=none seed=7; parms beta0 &init1 beta1 &init2; parms sigma2 &init3; prior beta0 beta1 ~ normal(0, var = 1e6); prior sigma2 ~ igamma(3/10, scale = 10/3); mu = beta0 + beta1*height;
Example 52.13: Gelman-Rubin Diagnostics
F
4319
model weight ~ normal(mu, var = sigma2); run;
%end;
%mend; ods listing close;
%gmcmc; ods listing;
The macro variables nchain
, nparm
, nsim
, and var define the number of chains, the number of parameters, the number of Markov chain simulations, and the parameter names, respectively. The macro GMCMC gets initial values from the data set init , assigns them to the macro variables init1
, init2 and init3
, starts the Markov chain at these initial values, and stores the posterior draws to three output data sets: out1 , out2 , and out3 .
In the PROC MCMC statement, the
option restarts the Markov chain after tuning at the assigned initial values. No burn-in is requested.
You can use the autocall macro GELMAN to calculate the Gelman-Rubin statistics by using the three chains. The GELMAN macro has the following arguments:
%macro gelman(dset, nparm, var, nsim, nc=3, alpha=0.05);
The argument dset is the name of the data set that stores the posterior samples from all the runs, nparm is the number of parameters, var is the name of the parameters, is the number of chains with a default value of 3, and alpha nsim is the ˛ is the number of simulations, nc significant level in the test with a default value of 0.05. This macro creates two data sets:
_Gelman_Ests and
_Gelman_Parms stores the names of the parameters.
stores the diagnostic estimates
The following statements calculate the Gelman-Rubin diagnostics:
data all; set out1(in=in1) out2(in=in2) out3(in=in3); if in1 then Chain=1; if in2 then Chain=2; if in3 then Chain=3; run;
%gelman(all, &nparm, &var, &nsim); data GelmanRubin(label='Gelman-Rubin Diagnostics'); merge _Gelman_Parms _Gelman_Ests; run; proc print data=GelmanRubin; run;
The Gelman-Rubin statistics are shown in
4320
F
Chapter 52: The MCMC Procedure
Output 52.13.1
Gelman-Rubin Diagnostics of the Regression Example
Obs
1
2
3
Simple Linear Regression, Gelman-Rubin Diagnostics
Parameter
Betweenchain
Withinchain Estimate
Upper
Bound beta0 beta1 sigma2
5384.76
1.20
8034.41
1168.64
0.30
2890.00
1.0002
1.0002
1.0010
1.0001
1.0002
1.0011
The Gelman-Rubin statistics do not reveal any concerns about the convergence or the mixing of the multiple chains. To get a better visual picture of the multiple chains, you can draw overlapping trace plots of these parameters from the three Markov chains runs.
The following statements create
/* plot the trace plots of three Markov chains. */
%macro trace;
%do i = 1 %to &nparm; proc sgplot data=all cycleattrs; series x=Iteration y=%scan(&var, &i) / group=Chain;
%end; run;
%mend;
%trace;
Output 52.13.2
Trace Plots of Three Chains for Each of the Parameters
Output 52.13.2
continued
Example 52.13: Gelman-Rubin Diagnostics
F
4321
4322
F
Chapter 52: The MCMC Procedure
The trace plots show that three chains all eventually converge to the same regions even though they started at very different locations. In addition to the trace plots, you can also plot the potential scale
reduction factor (PSRF). See the section “ Gelman and Rubin Diagnostics ” on page 160 for definition
and details.
The following statements calculate PSRF for each parameter. They use the GELMAN macro repeatedly and can take a while to run:
/* define sliding window size */
%let nwin = 200; data PSRF; run;
%macro PSRF(nsim);
%do k = 1 %to %sysevalf(&nsim/&nwin, floor);
%gelman(all, &nparm, &var, nsim=%sysevalf(&k*&nwin)); data GelmanRubin; run; merge _Gelman_Parms _Gelman_Ests; data PSRF; set PSRF GelmanRubin; run;
%end;
%mend PSRF; options nonotes;
%PSRF(&nsim); options notes; data PSRF; set PSRF; if _n_ = 1 then delete; run; proc sort data=PSRF; by Parameter; run;
%macro sepPSRF(nparm=, var=, nsim=);
%do k = 1 %to &nparm; data save&k; set PSRF; if _n_ > %sysevalf(&k*&nsim/&nwin, floor) then delete; if _n_ < %sysevalf((&k-1)*&nsim/&nwin + 1, floor) then delete;
Iteration + &nwin; run; proc sgplot data=save&k(firstobs=10) cycleattrs; series x=Iteration y=Estimate; series x=Iteration y=upperbound; yaxis label="%scan(&var, &k)"; run;
%end;
%mend sepPSRF;
%sepPSRF(nparm=&nparm, var=&var, nsim=&nsim);
Example 52.13: Gelman-Rubin Diagnostics
F
4323
Output 52.13.3
PSRF Plot for Each Parameter
4324
F
Chapter 52: The MCMC Procedure
Output 52.13.3
continued
PSRF is the square root of the ratio of the between-chain variance and the within-chain variance. A large PSRF indicates that the between-chain variance is substantially greater than the within-chain variance, so that longer simulation is needed. You want the PSRF to converge to 1 eventually, as it appears to be the case in this simulation study.
Aitkin, M., Anderson, D., Francis, B., and Hinde, J. (1989), Statistical Modelling in GLIM , Oxford:
Oxford Science Publications.
Atkinson, A. C. (1979), “The Computer Generation of Poisson Random Variables,” Applied Statistics ,
28, 29–35.
Atkinson, A. C. and Whittaker, J. (1976), “A Switching Algorithm for the Generation of Beta
Random Variables with at Least One Parameter Less Than One,” Proceedings of the Royal Society of London, Series A , 139, 462–467.
Bacon, D. W. and Watts, D. G. (1971), “Estimating the Transition between Two Intersecting Straight
Lines,” Biometrika , 58, 525–534.
Berger, J. O. (1985), Statistical Decision Theory and Bayesian Analysis , Second Edition, New York:
Springer-Verlag.
References
F
4325
Box, G. E. P. and Cox, D. R. (1964), “An Analysis of Transformations,”
Journal of the Royal
Statistics Society, Series B , 26, 211–234.
Carlin, B. P., Gelfand, A. E., and Smith, A. F. M. (1992), “Hierarchical Bayesian Analysis of
Changepoint Problems,” Applied Statistics , 41(2), 389–405.
Chaloner, K. (1994), “Residual Analysis and Outliers in Bayesian Hierarchical Models,” in Aspects of Uncertainty: A Tribute to D. V. Lindley , 149–157, New York: Wiley.
Chaloner, K. and Brant, R. (1988), “A Bayesian Approach to Outlier Detection and Residual
Analysis,” Biometrika , 75(4), 651–659.
Cheng, R. C. H. (1978), “Generating Beta Variates with Non-integral Shape Parameters,” Communications ACM
, 28, 290–295.
Congdon, P. (2003), Applied Bayesian Modeling , John Wiley & Sons.
Crowder, M. J. (1978), “Beta-Binomial Anova for Proportions,” Applied Statistics , 27, 34–37.
Draper, D. (1996), “Discussion of the Paper by Lee and Nelder,” Journal of the Royal Statistical
Society, Series B
, 58, 662–663.
Eilers, P. H. C. and Marx, B. D. (1996), “Flexible Smoothing with B -Splines and Penalties,” Statistical
Science , 11, 89–121, with discussion.
Finney, D. J. (1947), “The Estimation from Individual Records of the Relationship between Dose and Quantal Response,” Biometrika , 34, 320–334.
Fisher, R. A. (1935), “The Fiducial Argument in Statistical Inference,”
391–398.
Annals of Eugenics , 6,
Fishman, G. S. (1996), Monte Carlo: Concepts, Algorithms, and Applications , New York: John
Wiley & Sons.
Gaver, D. P. and O’Muircheartaigh, I. G. (1987), “Robust Empirical Bayes Analysis of Event Rates,”
Technometrics
, 29, 1–15.
Gelman, A., Carlin, J., Stern, H., and Rubin, D. (2004), Bayesian Data Analysis , Second Edition,
London: Chapman & Hall.
Gentleman, R. and Geyer, C. J. (1994), “Maximum Likelihood for Interval Censored Data: Consistency and Computation,” Biometrika , 81, 618–623.
Gilks, W. (2003), “Adaptive Metropolis Rejection Sampling (ARMS),” software from MRC
Biostatistics Unit, Cambridge, UK, http://www.maths.leeds.ac.uk/~wally.gilks/ adaptive.rejection/web_page/Welcome.html
.
Gilks, W. R. and Wild, P. (1992), “Adaptive Rejection Sampling for Gibbs Sampling,”
Statistics
, 41, 337–348.
Applied
Holmes, C. C. and Held, L. (2006), “Bayesian Auxiliary Variable Models for Binary and Multinomial Regression,” Bayesian Analysis , 1(1), 145–168,
2006/vol01/issue01/held.pdf
.
http://ba.stat.cmu.edu/journal/
4326
F
Chapter 52: The MCMC Procedure
Ibrahim, J. G., Chen, M. H., and Sinha, D. (2001),
Bayesian Survival Analysis
, New York: Springer-
Verlag.
Kass, R. E., Carlin, B. P., Gelman, A., and Neal, R. (1998), “Markov Chain Monte Carlo in Practice:
A Roundtable Discussion,” The American Statistician , 52, 93–100.
Krall, J. M., Uthoff, V. A., and Harley, J. B. (1975), “A Step-up Procedure for Selecting Variables
Associated with Survival,” Biometrics , 31, 49–57.
Kuhfeld, W. F. (2004), Conjoint Analysis , Technical report, SAS Institute Inc., http://support.
sas.com/resources/papers/tnote/tnote_marketresearch.html
.
Matsumoto, M. and Kurita, Y. (1992), “Twisted GFSR Generators,” ACM Transactions on Modeling and Computer Simulation
, 2(3), 179–194.
Matsumoto, M. and Kurita, Y. (1994), “Twisted GFSR Generators,” ACM Transactions on Modeling and Computer Simulation , 4(3), 254–266.
Matsumoto, M. and Nishimura, T. (1998), “Mersenne Twister: A 623-Dimensionally Equidistributed
Uniform Pseudo-Random Number Generator,” ACM Transactions on Modeling and Computer
Simulation , 8, 3–30.
McGrath, E. J. and Irving, D. C. (1973), Techniques for Efficient Monte Carlo Simulation, Volume
II: Random Number Generation for Selected Probability Distributions
, Technical report, Science
Applications Inc., La Jolla, CA.
Michael, J. R., Schucany, W. R., and Haas, R. W. (1976), “Generating Random Variates Using
Transformations with Multiple Roots,” American Statistician , 30(2), 88–90.
Pregibon, D. (1981), “Logistic Regression Diagnostics,” Annals of Statistics , 9, 705–724.
Ripley, B. D. (1987), Stochastic Simulation , New York: John Wiley & Sons.
Robert, C. (1995), “Simulation of Truncated Normal Variables,”
Statistics and Computing
, 5, 121–
125.
Roberts, G. O., Gelman, A., and Gilks, W. R. (1997), “Weak Convergence and Optimal Scaling of
Random Walk Metropolis Algorithms,” Annual of Applied Probability , 7, 110–120.
Roberts, G. O. and Rosenthal, J. S. (2001), “Optimal Scaling for Various Metropolis-Hastings
Algorithms,” Statistical Science , 16, 351–367.
Rubin, D. B. (1981), “Estimation in Parallel Randomized Experiments,” Journal of Educational
Statistics
, 6, 377–411.
Schervish, M. J. (1995), Theory of Statistics , New York: Springer-Verlag.
Sharples, L. (1990), “Identification and Accommodation of Outliers in General Hierarchical Models,”
Biometrika
, 77, 445–453.
Spiegelhalter, D. J., Thomas, A., Best, N. G., and Gilks, W. R. (1996), “BUGS Examples, Volume 2,
Version 0.5, (version ii),” .
arrays
MCMC procedure,
monitor values of (MCMC),
Behrens-Fisher problem
MCMC procedure,
Bernoulli distribution definition of (MCMC),
MCMC procedure,
beta distribution definition of (MCMC),
MCMC procedure,
binary distribution definition of (MCMC),
MCMC procedure,
binomial distribution definition of (MCMC),
MCMC procedure,
blocking
MCMC procedure,
Box-Cox transformation estimate
D
0 ,
MCMC procedure,
Cauchy distribution definition of (MCMC),
MCMC procedure,
censoring
MCMC procedure,
chi-square distribution definition of (MCMC),
MCMC procedure,
constants specification
MCMC procedure,
convergence
MCMC procedure,
Cox models
MCMC procedure,
dgeneral distribution
MCMC procedure,
dlogden distribution
MCMC procedure,
double exponential distribution definition of (MCMC),
MCMC procedure,
examples, MCMC array subscripts,
arrays,
arrays, store data set variables,
BEGINCNST/ENDCNST statements,
Behrens-Fisher problem,
blocking,
Box-Cox transformation,
censoring,
change point models,
cloglog transformation,
constrained analysis,
Cox models,
Cox models, time dependent covariates,
Cox models, time independent covariates,
deviance information criterion,
discrete priors,
error finding using the PUT statement,
estimate functionals,
estimate posterior probabilities,
exponential models, survival analysis,
FCMP procedure,
Gelman-Rubin diagnostics,
generalized linear models,
GENMOD procedure, BAYES statement,
getting started,
GLIMMIX procedure,
graphics, box plots,
graphics, custom template,
graphics, fit plots,
graphics, kernel density comparisons,
graphics, multiple chains,
graphics, posterior predictive checks,
graphics, PSRF plots,
graphics, scatter plots,
graphics, survival curves,
hierarchical centering,
IF-ELSE statement,
implement a conjugate Gibbs sampler,
implement a new sampling algorithm,
improve mixing,
improving mixing,
initial values,
interval censoring,
Jeffreys’ prior,
JOINTMODEL option,
LAG functions,
linear regression,
log transformation,
logistic regression, diffuse prior,
logistic regression, Jeffreys’ prior,
logistic regression, random-effects,
logistic regression, sampling via Gibbs,
logit transformation,
matrix functions,
MISSING= option,
mixed-effects models,
mixing,
mixture of normal densities,
model comparison,
modelling dependent data,
MONITOR= option, arrays,
multivariate priors,
NLMIXED procedure,
nonlinear Poisson regression,
PHREG procedure, BAYES statement,
Poisson regression,
Poisson regression, nonlinear,
Poisson regression, random-effects,
posterior predictive distribution,
probit transformation,
proportional hazard models,
random-effects models,
regenerate diagnostics plots,
SGPLOT procedure,
SGRENDER procedure,
specifying a new distribution,
store data set variables in arrays,
survival analysis,
survival analysis, exponential models,
survival analysis, Weibull model,
TEMPLATE procedure,
truncated distributions,
UDS statement,
use macros to construct loglikelihood,
user-defined samplers,
Weibull model, survival analysis,
exponential chi-square distribution definition of (MCMC),
MCMC procedure,
exponential distribution definition of (MCMC),
MCMC procedure,
exponential exponential distribution definition of (MCMC),
MCMC procedure,
exponential gamma distribution definition of (MCMC),
MCMC procedure,
exponential inverse chi-square distribution definition of (MCMC),
MCMC procedure,
exponential inverse-gamma distribution definition of (MCMC),
MCMC procedure,
exponential scaled inverse chi-square distribution definition of (MCMC),
MCMC procedure,
floating point errors
MCMC procedure,
gamma distribution definition of (MCMC),
MCMC procedure,
Gaussian distribution definition of (MCMC),
MCMC procedure,
Gelman-Rubin diagnostics
MCMC procedure,
general distribution
MCMC procedure,
generalized linear models
MCMC procedure,
geometric distribution definition of (MCMC),
MCMC procedure,
handling error messages
MCMC procedure,
hierarchical centering
MCMC procedure,
initial values
MCMC procedure,
inverse chi-square distribution definition of (MCMC),
MCMC procedure,
inverse Gaussian distribution definition of (MCMC),
MCMC procedure,
inverse-gamma distribution definition of (MCMC),
MCMC procedure,
Laplace distribution definition of (MCMC),
MCMC procedure,
likelihood function specification
MCMC procedure,
logden distribution
MCMC procedure,
logistic distribution definition of (MCMC),
MCMC procedure,
lognormal distribution definition of (MCMC),
MCMC procedure,
long run times
MCMC procedure,
marginal distribution
MCMC procedure,
Maximum a posteriori
MCMC procedure,
MCMC procedure,
arrays,
Behrens-Fisher problem,
Bernoulli distribution,
beta distribution,
binary distribution,
binomial distribution,
blocking,
Box-Cox transformation,
Cauchy distribution,
censoring,
chi-square distribution,
compared with other SAS procedures,
computational resources,
constants specification,
convergence,
Cox models,
deviance information criterion,
dgeneral distribution,
dlogden distribution,
double exponential distribution,
examples, see also examples, MCMC,
exponential chi-square distribution,
exponential distribution,
exponential exponential distribution,
exponential gamma distribution,
exponential inverse chi-square distribution,
exponential inverse-gamma distribution,
exponential scaled inverse chi-square distribution,
floating point errors,
gamma distribution,
Gaussian distribution,
Gelman-Rubin diagnostics,
general distribution,
generalized linear models,
geometric distribution,
handling error messages,
hierarchical centering,
hyperprior distribution,
initial values,
inverse chi-square distribution,
inverse Gaussian distribution,
inverse-gamma distribution,
Laplace distribution,
likelihood function specification,
logden distribution,
logistic distribution,
lognormal distribution,
long run times,
marginal distribution,
Maximum a posteriori,
mixed-effects models,
mixing,
model specification,
modeling dependent data,
negative binomial distribution,
nonlinear Poisson regression,
normal distribution,
options,
options summary,
output ODS Graphics table names,
output table names,
overflows,
parameters specification,
pareto distribution,
Poisson distribution,
posterior predictive distribution,
posterior samples data set,
precision of solution,
prior distribution,
prior predictive distribution,
programming statements,
proposal distribution,
random-effects models,
run times,
scaled inverse chi-square distribution,
specifying a new distribution,
standard distributions,
survival analysis,
syntax summary,
t distribution,
truncated distributions,
tuning,
UDS statement,
uniform distribution,
user defined sampler statement,
user-defined distribution,
user-defined samplers,
using the IF-ELSE logical control,
Wald distribution,
Weibull distribution,
mixed-effects models
MCMC procedure,
mixing convergence (MCMC),
improving (MCMC),
MCMC procedure,
model specification
MCMC procedure,
negative binomial distribution definition of (MCMC),
MCMC procedure,
nonlinear Poisson regression
MCMC procedure,
normal distribution definition of (MCMC),
MCMC procedure,
output ODS Graphics table names
MCMC procedure,
output table names
MCMC procedure,
overflows
MCMC procedure,
parameters specification
MCMC procedure,
pareto distribution definition of (MCMC),
MCMC procedure,
Poisson distribution definition of (MCMC),
MCMC procedure,
posterior predictive distribution
MCMC procedure,
precision of solution
MCMC procedure,
prior distribution data-set-dependent (MCMC),
distribution specification (MCMC),
hyperprior specification (MCMC),
predictive distribution (MCMC),
user-defined (MCMC),
programming statements
MCMC procedure,
proposal distribution
MCMC procedure,
random-effects models
MCMC procedure,
run times
MCMC procedure,
scaled inverse chi-square distribution definition of (MCMC),
MCMC procedure,
specifying a new distribution
MCMC procedure,
standard distributions
MCMC procedure,
survival analysis
MCMC procedure,
t distribution definition of (MCMC),
MCMC procedure,
truncated distributions
MCMC procedure,
tuning
MCMC procedure,
UDS statement
MCMC procedure,
uniform distribution definition of (MCMC),
MCMC procedure,
user defined sampler statement
MCMC procedure,
user-defined distribution
MCMC procedure,
user-defined samplers
MCMC procedure,
using the IF-ELSE logical control
MCMC procedure,
Wald distribution definition of (MCMC),
MCMC procedure,
Weibull distribution definition of (MCMC),
MCMC procedure,
ACCEPTTOL= option
PROC MCMC statement,
ARRAY statement
MCMC procedure,
AUTOCORLAG= option
PROC MCMC statement,
BEGINCNST statement
MCMC procedure,
BEGINNODATA statement
MCMC procedure,
BEGINPRIOR statement
MCMC procedure,
BY statement
PROC MCMC procedure,
COVARIATES= option
PREDDIST statement (MCMC),
DATA= option
PROC MCMC statement,
DIAG= option
PROC MCMC statement,
DIAGNOSTICS= option
PROC MCMC statement,
DIC option
PROC MCMC statement,
DISCRETE= option
PROC MCMC statement,
ENDCNST statement
MCMC procedure,
ENDNODATA statement
MCMC procedure,
ENDPRIOR statement
MCMC procedure,
HYPERPRIOR statement
MCMC procedure,
INF= option
PROC MCMC statement,
INIT= option
PROC MCMC statement,
JOINTMODEL option
PROC MCMC statement,
LIST option
PROC MCMC statement,
LISTCODE option
PROC MCMC statement,
MAXTUNE= option
PROC MCMC statement,
MCMC procedure,
ARRAY statement,
BEGINCNST statement,
BEGINNODATA statement,
BEGINPRIOR statement,
ENDCNST statement,
ENDNODATA statement,
ENDPRIOR statement,
HYPERPRIOR statement,
MODEL statement,
PARMS statement,
PRED statement,
PREDDIST statement,
PRIOR statement,
syntax,
MCMC procedure, ARRAY statement,
MCMC procedure, BEGINCNST statement,
MCMC procedure, BEGINNODATA statement,
MCMC procedure, BEGINPRIOR statement,
MCMC procedure, ENDCNST statement,
MCMC procedure, ENDNODATA statement,
MCMC procedure, ENDPRIOR statement,
MCMC procedure, HYPERPRIOR statement,
MCMC procedure, MODEL statement,
MCMC procedure, PARMS statement,
MCMC procedure, PRED statement,
MCMC procedure, PREDDIST statement,
COVARIATES= option,
NSIM= option,
OUTPRED= option,
STATISTICS= option,
STATS= option,
MCMC procedure, PRIOR statement,
MCMC procedure, PROC MCMC statement
ACCEPTTOL= option,
AUTOCORLAG= option,
DATA= option,
DIAG= option,
DIAGNOSTICS= option,
DIC option,
DISCRETE= option,
INF= option,
INIT= option,
JOINTMODEL option,
LIST option,
LISTCODE option,
MAXTUNE= option,
MINTUNE= option,
MISSING= option,
MONITOR= option,
NBI= option,
NMC= option,
NTU= option,
OUTPOST=option,
PLOTS= option,
PROPCOV= option,
PROPDIST= option,
SCALE option,
SEED option,
SIMREPORT= option,
SINGDEN= option,
STATISTICS= option,
STATS= option,
TARGACCEPT= option,
TARGACCEPTI= option,
THIN= option,
TRACE option,
TUNEWT= option,
MCMC procedure, Programming statements
ABORT statement,
CALL statement,
DELETE statement,
DO statement,
GOTO statement,
IF statement,
LINK statement,
PUT statement,
RETURN statement,
SELECT statement,
STOP statement,
SUBSTR statement,
WHEN statement,
MINTUNE= option
PROC MCMC statement,
MISSING= option
PROC MCMC statement,
MODEL statement
MCMC procedure,
MONITOR= option
PROC MCMC statement,
NBI= option
PROC MCMC statement,
NMC= option
PROC MCMC statement,
NSIM= option
PREDDIST statement (MCMC),
NTU= option
PROC MCMC statement,
OUTPOST= option
PROC MCMC statement,
OUTPRED= option
PREDDIST statement (MCMC),
PARMS statement
MCMC procedure,
PLOTS= option
PROC MCMC statement,
PRED statement
MCMC procedure,
PREDDIST statement
MCMC procedure,
PRIOR statement
MCMC procedure,
PROC MCMC procedure, BY statement,
PROPCOV=method
PROC MCMC statement,
PROPDIST= option
PROC MCMC statement,
SCALE option
PROC MCMC statement,
SEED option
PROC MCMC statement,
SIMREPORT= option
PROC MCMC statement,
SINGDEN= option
PROC MCMC statement,
STATISTICS= option
PREDDIST statement (MCMC),
PROC MCMC statement,
STATS= option
PREDDIST statement (MCMC),
PROC MCMC statement,
TARGACCEPT= option
PROC MCMC statement,
TARGACCEPTI= option
PROC MCMC statement,
THIN= option
PROC MCMC statement,
TRACE option
PROC MCMC statement,
TUNEWT= option
PROC MCMC statement,
We welcome your feedback.
If you have comments about this book, please send them to
. Include the full title and page numbers (if applicable).
If you have comments about the software, please send them to
.
®
®
®
Need to learn the basics? Struggling with a programming problem? You’ll find the expert answers that you need in example-rich books from SAS Press. Written by experienced SAS professionals from around the world, SAS Press books deliver real-world insights on a broad range of topics for all skill levels.
s u p p o r t . s a s . c o m / s a s p r e s s
®
To successfully implement applications using SAS software, companies in every industry and on every continent all turn to the one source for accurate, timely, and reliable information: SAS documentation.
We currently produce the following types of reference documentation to improve your work experience:
• Online help that is built into the software.
• Tutorials that are integrated into the product.
• Reference documentation delivered in HTML and PDF –
free
on the Web.
• Hard-copy books.
s u p p o r t . s a s . c o m / p u b l i s h i n g
®
Subscribe to SAS Publishing News to receive up-to-date information about all new SAS titles, author podcasts, and new Web site features via e-mail. Complete instructions on how to subscribe, as well as access to past issues, are available at our Web site.
s u p p o r t . s a s . c o m / s p n
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.
Other brand and product names are trademarks of their respective companies. © 2009 SAS Institute Inc. All rights reserved. 518177_1US.0109
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
© 2021