CEMDAP CEMDAP v 3.1, CEMDAP 10Threads, CEMDAP 5Threads Simulation Software User's Manual

CEMDAP
User’s Manual
Home
12:30 PM
7:30 AM
Pers.
8:20 AM
Buss.
9:16 AM
11:37 AM
Pers. Buss.
9:40 AM
CEMDAP v 3.1 User’s Manual
Table of Contents
TABLE OF CONTENTS .......................................................................................................................... 1
INTRODUCTION ....................................................................................................................................... 1
CHAPTER 1. SYSTEM REQUIREMENTS AND INSTALLATION INSTRUCTIONS ........... 3
1.1 MINIMUM SYSTEM REQUIREMENTS .................................................................................... 3
1.2 INSTALLATION INSTRUCTIONS............................................................................................... 3
1.3 TO UNINSTALL .............................................................................................................................. 6
CHAPTER 2. THE CEMDAP ENVIRONMENT ............................................................................... 9
CHAPTER 3. INPUT AND OUTPUT FILES ................................................................................... 13
3.1 INPUT DATA SPECIFICATION ................................................................................................. 13
3.1.1 Data Inputs ........................................................................................................................... 13
3.1.2 Model Specifications........................................................................................................... 15
3.2 OUTPUT DATA FILES ................................................................................................................ 16
CHAPTER 4. BASIC OPERATIONS ................................................................................................. 19
4.1 REGISTERING THE INPUT DATABASE ............................................................................... 19
4.2 LOADING THE INPUT DATA FILE .......................................................................................... 23
4.3 SPECIFYING THE OUTPUT FILES ......................................................................................... 24
4.4 CONFIGURING THE MODEL SYSTEM ................................................................................. 25
4.4.1 Accessing model module dialog boxes through the Models Menu......................... 25
4.4.2 Accessing model module dialog boxes through the Interactive User Interface .. 31
4.4.3 Saving model configuration file ....................................................................................... 44
4.4.4 Loading model configuration file..................................................................................... 46
4.4.5 Configuring specific model types .................................................................................... 47
4.5 INITIATING A SIMULATION RUN ........................................................................................... 58
REFERENCES ......................................................................................................................................... 61
APPENDIX A. D-FW INPUT DATA ....................................................................................................... 63
APPENDIX B. INPUT DATABASE PREPARATION IN POSTGRESQL FORMAT ....................... 69
APPENDIX C. STRUCTURE OF THE MODEL SPECIFICATION FILE ........................................ 83
CEMDAP v 3.1 User’s Manual
CEMDAP v 3.1 User’s Manual
List of Figures
Figure 1.1: Cemdap Setup Wizard Welcome Screen.......................................................... 3
Figure 1.2: Select Installation Folder Screen ...................................................................... 4
Figure 1.3: Confirm Installation Screen.............................................................................. 5
Figure 1.4: Installation Complete Screen............................................................................ 6
Figure 1.5: Cemdap Setup Wizard Welcome Screen once CEMDAP is Installed ............. 7
Figure 1.6: Installation is Complete Screen once CEMDAP is removed ........................... 7
Figure 2.1: CEMDAP Main Window and Menus .............................................................. 9
Figure 4.1: Input command in the Data Menu .................................................................. 19
Figure 4.2: Machine Data Source Tab within the Select Data Source Screen .................. 20
Figure 4.3: Create New Data Source Dialog Box ............................................................. 20
Figure 4.4: Select driver Screen within Create New Data Source Dialog Box ................ 21
Figure 4.5: Create New Data Source Finish Screen ......................................................... 22
Figure 4.6: PostgreSQL ANSI ODBC Dialog Box .......................................................... 22
Figure 4.7: Final Select Data Source Screen .................................................................... 23
Figure 4.8: Input command in the Data menu (2) ............................................................. 24
Figure 4.9: Output command in the Data menu ................................................................ 24
Figure 4.10: Save Simulation Results As… dialog box ................................................... 25
Figure 4.11: List of Work and School Activities Models within Generation
Allocation in the Models menu ............................................................................. 26
Figure 4.12: List of Children’s Activities Models within Generation Allocation in
the Models menu ................................................................................................... 26
Figure 4.13: List of Independent Activities Models within Generation Allocation
in the Models menu ............................................................................................... 27
Figure 4.14: List of Scheduling – Children Models within the Models menu ................. 27
Figure 4.15: List of Scheduling – Worker Models within the Models menu ................... 28
Figure 4.16: List of Number of Tours Models within Scheduling-Worker in the
Models menu ......................................................................................................... 28
Figure 4.17: List of Activity Type at Stop Models within Scheduling-Worker in
the Models menu ................................................................................................... 29
Figure 4.18: List of Scheduling – Non Worker Models within the Models menu ........... 29
Figure 4.19: List of Tour Mode Models within Scheduling – Non Worker in the
Models menu ......................................................................................................... 30
Figure 4.20: List of Activity Type at Stop Models within Scheduling – Non
Worker in the Models menu ................................................................................. 30
Figure 4.21: List of Departure Time from Home Models within Scheduling –
Joint in the Models menu ..................................................................................... 31
Figure 4.22: Interactive UI in the Models menu ............................................................... 31
Figure 4.23: Main Dialog Box of the Interactive User Interface ...................................... 32
Figure 4.24: Generation-Allocation Model System Interactive User Interface ................ 33
Figure 4.25: Generation & Timing of Work and School Activities of Adults
Interactive User Interface ...................................................................................... 34
CEMDAP v 3.1 User’s Manual
Figure 4.26: Generation of Children’s Travel Needs and Allocation of
Responsibilities to Parents Interactive User Interface .......................................... 35
Figure 4.27: Generation of Independent Activities by Adults and Children for
Personal and Household Serving Needs Interactive User Interface ..................... 36
Figure 4.28: Scheduling Model System Interactive User Interface .................................. 37
Figure 4.29: Scheduling of Work to Home Commute for Worker Interactive User
Interface ................................................................................................................ 38
Figure 4.30: Scheduling of the Home to Work Commute for a Worker Interactive
User Interface ........................................................................................................ 39
Figure 4.31: Scheduling of the Tour Containing the Drop-off Activity for a
Nonworker Interactive User Interface .................................................................. 40
Figure 4.32: Scheduling of the Tour Containing the Pick-up Activity for a
Nonworker Interactive User Interface .................................................................. 41
Figure 4.33: Scheduling of the Tour Containing the Pick-up Activity for a
Nonworker Interactive User Interface .................................................................. 41
Figure 4.34: Scheduling of the Joint Discretionary Tour Undertaken by a Worker
with Children Interactive User Interface............................................................... 42
Figure 4.35: Scheduling of the Joint Discretionary Tour Undertaken by a
Nonworker with Children Interactive User Interface ........................................... 42
Figure 4.36: Scheduling of the Independent Home-based & Work-based Tours for
Workers… Interactive User Interface ................................................................... 43
Figure 4.37: Scheduling Independent Tours for Nonworkers Interactive User
Interface ................................................................................................................ 43
Figure 4.38: Scheduling the Independent Discretionary Tour Undertaken by a
Child Interactive User Interface ............................................................................ 44
Figure 4.39: Save command in the Models Menu ............................................................ 45
Figure 4.40: Save As Dialog Box from the Save Command ............................................ 45
Figure 4.41: Load Command in the Models Menu ........................................................... 46
Figure 4.42: Save As Dialog Box from the Load Command ............................................ 47
Figure 4.43: Linear Regression Model Interactive User Interface ................................... 48
Figure 4.44: Example of Fully Configured Linear Regression Model ............................. 49
Figure 4.45: Hazard Duration Model Interactive User Interface ...................................... 50
Figure 4.46: Example of Fully Configured Hazard Duration Model................................ 51
Figure 4.47: Multinomial Logit Model Interactive User Interface ................................... 52
Figure 4.48: Example of Constant and Corresponding Coefficient for the Utility
Function of one of the Multinomial Logit Model Alternatives ............................ 53
Figure 4.49: Example of Fully Configured Multinomial Logit Model ............................. 54
Figure 4.50: Binary Logit Model Interactive User Interface ............................................ 55
Figure 4.51: Example of Fully Configured Binary Logit Model ...................................... 56
Figure 4.52: Example of Fully Configured Ordered Probit Model .................................. 57
Figure 4.53: Work-Start-End-Time Model Interactive User Interface ............................. 58
Figure 4.54: Run command in the Simulation Menu........................................................ 58
Figure 4.55: Simulation in Progress Dialog Box .............................................................. 59
Figure 4.56: Simulation Completed Successfully Dialog Box ......................................... 59
CEMDAP v 3.1 User’s Manual
Figure B-1 Run Command in Start menu of Windows..................................................... 76
Figure B-2 Run dialog Box ............................................................................................... 76
Figure B-3 Command Prompt dialog box: cd\ .................................................................. 77
Figure B-4 Command Prompt dialog box: cd Program files\postgresql\8.2\bin .............. 77
Figure B-5 Command Prompt dialog box: psql –d database name –U user name .......... 78
Figure B-6 Command Prompt dialog box: C:\\Program
Files\\PostgreSQL\\8.2\\bin\\TestData\\households.dat........................................ 79
CEMDAP v 3.1 User’s Manual
CEMDAP v 3.1 User’s Manual
List of Tables
Table 2.1: Data Menu ....................................................................................................... 10
Table 2.2: Models Menu ................................................................................................... 10
Table 2.3: Simulation Menu ............................................................................................. 11
Table 2.4: Help Menu ....................................................................................................... 11
Table 3.1: Households Table with Required Variables .................................................... 13
Table 3.2: Persons Table with Required Variables ........................................................... 14
Table 3.3: Zones Table with Required Variables ............................................................. 14
Table 3.4: Zone2Zone Table and Required Variables ...................................................... 14
Table 3.5: LOSdir Table with Required Variables ........................................................... 14
Table 3.6: LOS Table with Required Variables................................................................ 15
Table 3.7: Adults.Out Format and Content ....................................................................... 16
Table 3.8: Children.Out Format and Content ................................................................... 17
Table 3.9: Workers.Out Format and Content.................................................................... 17
Table 3.10: Childstu.Out Format and Content .................................................................. 17
Table 3.11: Nonworkers.Out Format and Content............................................................ 17
Table 3.12: Tours.Out Format and Content ...................................................................... 18
Table 3.13: Stops.Out Format and Content ...................................................................... 18
Table A.1 Households Table Specification for D-FW Area ............................................. 63
Table A.2 Persons Table Specification for D-FW Area ................................................... 64
Table A.3 Zones Table Specification for DF-W Area ...................................................... 65
Table A.4 Zone2Zone Table Specification for DF-W Area ............................................. 65
Table A.5 LOS Tables (AMPeak, PMPeak, and OFFPeak) Specifications for DFW Area .................................................................................................................. 66
Table A.6 LOSDir Tables (AMPeak, PMPeak, and OFFPeak) Specifications for
DF-W Area............................................................................................................ 66
Table A.7 LOSdir Table Currently Present in CEMDAP................................................. 66
Table B.1 Column names of Postgres Tables ................................................................... 81
CEMDAP v 3.1 User’s Manual
CEMDAP v 3.1 User’s Manual
Introduction
The “Comprehensive Econometric Micro-simulator for Daily Activity-Travel
Patterns” (CEMDAP) is a software implementation of a system of econometric models
that represent the decision-making behavior of individuals. It is one of the first systems to
comprehensively simulate the activity-travel patterns of all household members (i.e., both
adults and children, further distinguished based on employment/student status) in a
continuous time domain, while also accommodating interactions among household
members.
Given various land-use, socio-demographic, activity system, and
transportation level-of-service attributes as input, the system provides as output the
complete daily activity-travel patterns for each individual in a population.
CEMDAP has been developed using an Object Oriented Design. It was
implemented in Visual C++ using the Visual Studio .NET development platform. The
software provides a user-friendly Windows environment to configure or modify the
components of the modeling system, save or load model configurations, load the input
data, and run the simulation. For an overview of the modeling system embedded in
CEMDAP, the reader is referred to the Guo et al. (2005) and Pinjari et al. (2006).
The rest of this manual is organized as follows. Chapter 1 describes the system
requirements and installation instructions for the software. Chapter 2 provides an
overview of the CEMDAP environment, including a description of the menus and
commands available. Chapter 3 describes the prescribed input and output file
specifications and formats. Chapter 4 guides the user through the basic operations such as
loading inputs, and configuring model components.
1
CEMDAP v 3.1 User’s Manual
2
CEMDAP v 3.1 User’s Manual
Chapter 1. Installation Instructions
1.1 MINIMUM SYSTEM REQUIREMENTS
CEMDAP requires a Pentium IV-class processor with 1 Giga byte (or above) of
Random Access Memory (RAM) and 2 Giga Hertz (or above) processing speed. The
operating systems supported include Windows NT 4, Windows 2000, and Windows XP.
1.2 INSTALLATION INSTRUCTIONS
To install CEMDAP in your system, find the Cemdap_5Threads.msi or
Cemdap_10Threads.msi1 file in the installation CD and follow the steps outlined below.
1. Double click Cemdap_5Threads.msi or Cemdap_10Threads.msi to run the
CEMDAP setup wizard.
2. In the Welcome screen, click Next to continue.
Figure 1.1: Cemdap Setup Wizard Welcome Screen
1
Use Cemdap_5Threads.msi for machines with Random Access Memory less than 1 GB, and use
Cemdap_10Threads.msi for machines with Random Access Memory greater than 1 GB.
3
CEMDAP v 3.1 User’s Manual
3. This leads to the Select Installation Folder screen. Choose the destination
location and folder name for the program installation (for example, C:\Program
Files\University of Texas\Cemdap) and click Next to continue.
Figure 1.2: Select Installation Folder Screen
4. Click Next in the Confirm Installation screen.
4
CEMDAP v 3.1 User’s Manual
Figure 1.3: Confirm Installation Screen
5. Installation will complete in a few seconds.
6. Click Close to exit the wizard.
5
CEMDAP v 3.1 User’s Manual
Figure 1.4: Installation Complete Screen
At the end of the installation process, CEMDAP will be added to the Program
menu, and a shortcut to the program will be created on the Desktop. Double click the
desktop shortcut to open the CEMDAP_5Threads/CEMDAP_10Threads folder. In this
folder, double click on cemdap.exe to start CEMDAP.
The installation CD contains sample input data files (in the TestData folder), and
the model specification (or the configuration) file for the Dallas-Fort-Worth area
implementation (further instructions to run CEMDAP are provided in Chapter 4). The
installation CD also contains an EXCEL files describing the CEMDAP inputs (input file
description) and outputs (output file description). Prior to running CEMDAP, it is
necessary to install PostgreSQL. The corresponding installation instructions are provided
in Appendix B on page 60.
1.3 TO UNINSTALL
To uninstall CEMDAP use the Cemdap_5Threads.msi or Cemdap_10Threads.msi
file in the installation CD and follow the instructions given below.
1. Double click Cemdap_5Threads.msi or Cemdap_10Threads.msi to run the
CEMDAP setup wizard.
2. In the Welcome screen select the Remove Cemdap radial button and click
Finish.
6
CEMDAP v 3.1 User’s Manual
Figure 1.5: Cemdap Setup Wizard Welcome Screen once CEMDAP is Installed
3. Click Close to exit.
Figure 1.6: Installation is Complete Screen once CEMDAP is removed
7
CEMDAP v 3.1 User’s Manual
8
CEMDAP v 3.1 User’s Manual
Chapter 2. The CEMDAP Environment
CEMDAP offers a user-friendly environment to simulate the activity-travel patterns
of a population by using standard Windows user interface features. In this chapter, we
present the basic features of the CEMDAP environment.
To start CEMDAP, double-click the shortcut to CEMDAP_5Threads or
CEMDAP_10Threads on your desktop, or find CEMDAP_5Threads or
CEMDAP_10Threads in the Start menu under All Programs and double click on
cemdap.exe. This opens up the following main window of CEMDAP.
Menus
Figure 2.1: CEMDAP Main Window and Menus
Main Window
The above main CEMDAP window remains open as long as the software is being
used, and it hosts the important menu items and other dialog boxes. The primary
functionality of CEMDAP lies within the menu commands. The commands below,
available from the menu bar, and their functions, are tabulated below.
9
CEMDAP v 3.1 User’s Manual
Table 2.1: Data Menu
Choose…
To…
Input
Load the input data. The input data must be ‘registered’ prior to
loading it (refer to Chapter 3 for details).
Output
Specify the names and locations for the output files. The output
file formats are described in detail in Chapter 3.
Exit
Quit from the CEMDAP environment.
Table 2.2: Models Menu
Choose…
To…
Load
Load the model configurations from file (refer to Chapter 4 on
Basic Operations for instructions on configuring models).
Save
Save the model configurations to file (refer to Chapter 4 on
Basic Operations for instructions on configuring models).
Generation Allocation
To access the model configuration dialog boxes for the suite of
generation-allocation model components (refer to Chapter 4).
Scheduling-Children
To access the model configuration dialog boxes for the suite of
Children-Scheduling model components (refer to Chapter 4).
Scheduling-Worker
To access the model configuration dialog boxes for the suite of
Worker-Scheduling model components (refer to Chapter 4).
Scheduling-Non Worker
To access the model configuration dialog boxes for the suite of
Non Worker-Scheduling model components (refer to Chapter
4).
Scheduling-Joint
To access the model configuration dialog boxes for the suite of
Joint Activity-Scheduling model components (refer to Chapter
4).
Interactive UI
To open the interactive user interface that ties all model
configuration dialog boxes together using the modeling system
framework (refer to Chapter 4).
10
CEMDAP v 3.1 User’s Manual
Table 2.3: Simulation Menu
Choose…
To…
Run
Initiate a simulation run. (Note that the input data must be
loaded and the complete model system configured prior to
using the Run command).
Table 2.4: Help Menu
Choose…
To…
About
Display the CEMDAP dialog box that shows copyright and
version information
11
CEMDAP v 3.1 User’s Manual
12
CEMDAP v 3.1 User’s Manual
Chapter 3. Input and Output Files
3.1 INPUT DATA SPECIFICATION
The inputs required by CEMDAP can be broadly classified into two categories: a)
data inputs, which include the population characteristics, zonal descriptives, and levelof-service data of the transportation network; and b) model parameters for all the
components of the embedded model system.
3.1.1 Data Inputs
The data inputs to CEMDAP are provided in the form of tables. There are six main
types of tables. The Household table contains the household-level characteristics (such as
residential location and number of vehicles) of the population, and the Persons table
contains the person-level characteristics (such as employment status and work location).
The Zones table comprises data on the characteristics of each of the traffic analysis zones
(such as population and employment). The Zone2Zone table includes data such as
distance between each of the zonal pairs. CEMDAP allows the analyst to specify the
inter-zonal level-of-service measures for any number of time periods. The Losdir table
identifies the names of the LOS tables included, and the time-of-day period
corresponding to each of these tables. Finally, there are as many LOS tables (with names
as specified in the Losdir table), each containing several inter-zonal LOS measures such
as travel times and costs for different modes.
Each table comprises several “required” data items or columns. These data include
record identifiers (such as household ID and person ID), and other basic information
required by the CEMDAP modeling system. In addition, the Households and Persons
tables may also have additional columns corresponding to additional explanatory
variables used in the underlying empirical models. Overall, each exogenous variable must
have a corresponding column in the appropriate table. The structure of each of the input
tables is presented below, identifying only the required data items and their ordering
within the tables. Additional columns are simply appended to these tables.
Table 3.1: Households Table with Required Variables
Column
1
2
3
4
5
6
Variable
Household ID
Number of adults
Total number of HH vehicles
Home TSZ location
Number of children
Household structure
13
Variable Name
HID
N_ADULTS
N_AUTOS
ZONE_ID
KIDS
Structure
CEMDAP v 3.1 User’s Manual
Table 3.2: Persons Table with Required Variables
Column
1
2
3
4
5
6
7
8
9
Variable
Household ID
Person ID
Employed
Student
Driver's license
Location of work
Location of school
Female
Parent
Variable Name
HID
PID
EMPLOYED
STUDYING
LICENSE
WORK_ZON
STUD_ZON
FEMALE
PARENT
Table 3.3: Zones Table with Required Variables
Column
1
2
3
4
5
6
7
8
9
10
Variable
Zone ID
Dallas CBD
Fort Worth CBD
Distance to the nearest shopping mall
Accessibility to retail employment
Accessibility to retail and service employment
Distance to the nearest shopping mall
Zonal service employment
Zonal retail employment
Zonal population
Variable Name
ZID
DALCBD
FWCBD
SHOPDIST
REMPACC
RSEMPACC
SHOPDIST
SEMP
REMP
NUMPERS
Table 3.4: Zone2Zone Table and Required Variables
Column
1
2
3
4
Variable
Origin Zone
Destination Zone
Adjacent
Distance
Variable Name
ORIG_ZON
DEST_ZON
ADJACENT
DISTANCE
Comment
1 if origin and destination zones are adjacent
Miles
Table 3.5: LOSdir Table with Required Variables
Column
1
2
3
Variable
Start time
End time
Table name
Variable Name
STAT_T
END_T
TBNAME
14
Comment
Start time of the time period of the LOS table
End time of the time period of the LOS table
Name (string) of the LOS table
CEMDAP v 3.1 User’s Manual
Table 3.6: LOS Table with Required Variables
Column
1
2
3
4
5
6
7
8
9
10
11
12
Variable
Origin zone
Destination zone
Same zone
Adjacent
Distance
DA IVTT
DA OVTT
Transit availability
Transit IVTT
Transit OVTT
Transit cost
DA cost
Variable Name
ORIG_ZON
DEST_ZON
SAME_ZON
ADJACENT
DISTANCE
DA_IVTT
DA_OVTT
TR_AVAIL
TR_IVTT
TR_OVTT
TR_COST
AU_COST
Comment
miles
minutes
minutes
minutes
minutes
Dollars
Dollars
Appendix A shows all the tables and variables, including the ‘required variables’
and some ‘additional variables’ used in the current implementation of CEMDAP for the
DFW area. The Data Inputs are required to be in the postgreSQL database format. The
input tables in the postgreSQL database must take titles as specified above and the
variables must be of the ‘double’ type. The variables names are not pre-specified; each
variable is identified by the specific table (table number) it belongs to, and the position
(column number) of the variable in the table. Appendix B presents the procedure to
prepare the input database corresponding to the DFW implementation in the postgreSQL
format.
3.1.2 Model Specifications
The model system embedded in CEMDAP is described in greater detail in Guo et
al. (2005) and Pinjari et al. (2006). The reader is referred to Pinjari et al. (2006) for a list
of all the model components. The variables and parameters of each of the model
components that constitute this model system must be specified the first time CEMDAP
is used. Thereafter, the model configurations can be saved into a file using the Save
command in the Models Menu and reloaded as desired using the Load command in the
Models Menu. The first time specification of the model configurations can be achieved in
one of the following four ways (refer to Section 4.4 for further details).
1. Access each of the model configuration dialog boxes through the commands in the
Models Menu. In particular, access the dialog boxes under the Generation Allocation,
Scheduling-Children, Scheduling-Worker, Scheduling-Non Worker, and SchedulingJoint entries in the Models Menu. (This method is recommended for changing the
parameters of a few models).
2. Access the model configuration dialog boxes through the Interactive UI, which is also
in the Models Menu. The interactive user interface ties all the dialog boxes together
15
CEMDAP v 3.1 User’s Manual
using the embedded modeling framework. (This method is recommended for first
time model parameter input).
3. The model parameters for all the model components can be coded into a text file
(model configuration file) in a prescribed format, and then loaded into the system
using the Load command in the Models Menu. A sample model configuration file can
be found in /data under the installation destination directory. Appendix C describes
the prescribed format of the model configuration file (For beginners and /or initial
users, it is recommended to specify the model parameters using the Graphical User
Interface instead of coding them in the model configuration file).
3.2 OUTPUT DATA FILES
CEMDAP produces as output the complete activity-travel patterns for a day for
every individual in the population of interest. There are seven output files, corresponding
to characteristics of (1) Adults, (2) Children, (3) Workers, (4) Students, (5) Non-workers,
(6) Tours, and (7) Stops. Each of these files is partitioned into smaller subsets based on
the number of threads. The files Adults.out and Children.out contain the decisions to
undertake activities of different types for adults and children, respectively. The files
Workers.out, Nonworkers.out, and Childstu.out contain the pattern-level attributes of the
workers’ (including adult students), non-workers’, and child students’ patterns,
respectively. The files Tours.out and Stops.out contain the tour-level and stop-level
attributes, respectively, for all the individuals that undertake travel. These output files can
easily be imported into any spreadsheet of your choice. The formats and content of these
output files are given below. Detailed description of the output variables can be found in
an EXCEL file (output file specification) in the /data under the installation destination
directory.
Table 3.7: Adults.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Adult goes to work on the day
4
Adult undertakes work-related activity
5
Adult drops-off children at school
6
Adult picks-up children from school
7
Adult undertakes joint discretionary activities with children
8
Adult undertakes shopping activity
9
Adult undertakes HH/personal business activity
10
Adult undertakes social/recreational activity
11
Adult undertakes eat-out activity
12
Adult undertakes other serve passenger activity
16
CEMDAP v 3.1 User’s Manual
Table 3.8: Children.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Child goes to school on the day
4
Child undertakes joint discretionary activities with parent
5
Child undertakes independent discretionary activities
Table 3.9: Workers.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Work/school start time
4
Work/school end time
5
Number of before-work tours
6
Number of work-based tours
7
Number of after-work tours
Table 3.10: Childstu.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
School start time
4
School end time
5
Child gets dropped off at school by parent
6
Child gets picked up from school by parent
Table 3.11: Nonworkers.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Total number of tours made
17
CEMDAP v 3.1 User’s Manual
Table 3.12: Tours.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Tour identification number
4
Home/work stay duration before tour
5
Tour mode
6
Tour duration
7
Number of stops in tour
Table 3.13: Stops.Out Format and Content
Column No. Description
1
Household identification number
2
Person identification number
3
Tour identification number
4
Stop identification number
5
Activity type
6
Start time of travel to the stop
7
Travel time to stop
8
Stop duration
9
Stop location (zone) ID
10
Origin zone ID
11
Trip distance (zone to zone)
12
Activity type at the previous stop
18
CEMDAP v 3.1 User’s Manual
Chapter 4. Basic Operations
4.1 REGISTERING THE INPUT DATABASE
The input database, which is in the postgreSQL database format, must be registered
before it can be loaded in CEMDAP. The database registration should be done each time
a new database is used. PostgreSQL should be installed in the computer prior to
running CEMDAP. Further, data needs to be populated into the PostgreSQL database
(the reader is referred to Appendix B for installation and database loading instructions).
After installation of PostgreSQL and populating the database, follow the instructions
below to register an input file.
1. Click on the Input command in the Data Menu of CEMDAP.
Figure 4.1: Input command in the Data Menu
2. In the Select Data Source dialog that opens up, select the Machine Data Source
tab and click on the New button.
19
CEMDAP v 3.1 User’s Manual
Figure 4.2: Machine Data Source Tab within the Select Data Source Screen
3. In the Create New Data Source dialog that opens up, select the System Data
Source button and click on the Next button.
Figure 4.3: Create New Data Source Dialog Box
20
CEMDAP v 3.1 User’s Manual
4. Scroll down and select the PostgreSQL ANSI driver from the list of drivers and
click Next as displayed below.
Figure 4.4: Select driver Screen within Create New Data Source Dialog Box
5. Once you click Next, the following box will appear. Then, click Finish to open
up the PostgreSQL ANSI ODBC dialog box.
21
CEMDAP v 3.1 User’s Manual
Figure 4.5: Create New Data Source Finish Screen
6. In the PostgreSQL ANSI ODBC dialog box (shown below), enter the name of
the Data Source (Note The Data Source can be given any name, therefore you
can leave the default name), name of the Database (ex: CEMDAP_Test_Data.
Note this name should be the same as the database name that was created in data
loading process described in Appendix B), localhost as the name of the server,
the appropriate username and password of the postgreSQL data base installed in
the machine (see Appendix B), 5432 as the Port, prefer as the SSL Mode, and
provide a description (ex: Test Data) of the database. After entering all the
above mentioned items, click Save.
Figure 4.6: PostgreSQL ANSI ODBC Dialog Box
22
CEMDAP v 3.1 User’s Manual
7. Finalize the database registration process by clicking OK to close the dialog
boxes. CEMDAP is now set up to access the database file you registered.
Figure 4.7: Final Select Data Source Screen
4.2 LOADING THE INPUT DATA FILE
Any PostgreSQL database that satisfies the prescribed format (refer to Chapter 3),
and that has been registered as explained above, can be loaded in CEMDAP as input. The
following procedure should be followed to load the input data.
Note: If the database has just been registered, this step (i.e., loading the input
data file step) must be skipped. These steps are essential for loading an already
registered database.
1. Click on the Input command in the Data Menu of CEMDAP.
23
CEMDAP v 3.1 User’s Manual
Figure 4.8: Input command in the Data menu (2)
2. In the Select Data Source dialog that opens up, select the Machine Data Source
tab (first figure on this page), and choose the registered data source to be loaded
(for example, PostgreSQL30).
3. Click OK and the input data is loaded.
4.3 SPECIFYING THE OUTPUT FILES
During a simulation run, CEMDAP creates seven output files (refer to Chapter 3 for
file formats). The following steps are used to specify the location of these seven files.
1. Click on the Output command in the Data menu of CEMDAP.
Figure 4.9: Output command in the Data menu
2. In the Save Simulation Results As… dialog box that opens, click on Browse to
select the folder location for all the output files.
24
CEMDAP v 3.1 User’s Manual
Figure 4.10: Save Simulation Results As… dialog box
3. Select the location (or enter the path name) for all the output files, and then
click OK.
4.4 CONFIGURING THE MODEL SYSTEM
All the components of the model system embedded in CEMDAP must be
configured into a model configuration file (also referred to as a model specification file)
and input to the system before starting a simulation run. Model configuration involves
selecting the relevant variables and inputting their parameters. The model system can be
completely configured either by accessing all the corresponding dialog boxes to specify
the models or by manually preparing a model configuration file (as described in
Appendix C). Sections 4.4.1 and 4.4.2 describe two methods of accessing all the model
dialog boxes to configure the entire model system. Section 4.4.3 explains how the
complete model configuration entered using the dialog boxes can be saved in a file.
Section 4.4.4 explains how to load a model configuration file into CEMDAP. All the
model components fall into one of seven categories: Linear Regression, Hazard Duration,
Multinomial Logit, Binary Logit, Location Choice, Ordered Probit or Work Start/End
Time. Section 4.4.5 takes an example of each of these types to explain how they must be
configured.
Note: For Testing purposes, the reader is recommended to skip to Section 4.4.4
to load the ready-made model configuration file provided in the CEMDAP CD
(labeled as “CEMDAP_DFW_Model_Specification_file.cml”).
Menu
4.4.1 Accessing model module dialog boxes through the Models
The model configuration dialog boxes for each of the model components can be
accessed through the Models Menu. Each entry in the Models Menu allows access to a
specific suite of model components. Clicking on any of the models in each suite opens up
the corresponding dialog box to configure the model (see the following figures). For
example, the figure below shows that the models (such as “Children’s decision to go to
school”) can all be accessed by clicking on “Work and School Activities” under
25
CEMDAP v 3.1 User’s Manual
“Generation Allocation”. The subsequent figures (in the following pages) show how
each of the remaining CEMDAP model components can be directly accessed from the
menu items.
Figure 4.11: List of Work and School Activities Models within Generation Allocation in the
Models menu
Figure 4.12: List of Children’s Activities Models within Generation Allocation in the Models
menu
26
CEMDAP v 3.1 User’s Manual
Figure 4.13: List of Independent Activities Models within Generation Allocation in the Models
menu
Figure 4.14: List of Scheduling – Children Models within the Models menu
27
CEMDAP v 3.1 User’s Manual
Figure 4.15: List of Scheduling – Worker Models within the Models menu
Figure 4.16: List of Number of Tours Models within Scheduling-Worker in the Models menu
28
CEMDAP v 3.1 User’s Manual
Figure 4.17: List of Activity Type at Stop Models within Scheduling-Worker in the Models menu
Figure 4.18: List of Scheduling – Non Worker Models within the Models menu
29
CEMDAP v 3.1 User’s Manual
Figure 4.19: List of Tour Mode Models within Scheduling – Non Worker in the Models menu
Figure 4.20: List of Activity Type at Stop Models within Scheduling – Non Worker in the Models
menu
30
CEMDAP v 3.1 User’s Manual
Figure 4.21: List of Departure Time from Home Models within Scheduling – Joint in the
Models menu
4.4.2 Accessing model module dialog boxes through the
Interactive User Interface
Figure 4.22: Interactive UI in the Models menu
31
CEMDAP v 3.1 User’s Manual
The Interactive UI command in the Models Menu opens up the interactive user
interface that ties together the model configuration dialog boxes for all the model
components that constitute the embedded model system. This interactive UI is also
reflective of the underlying modeling framework embedded within CEMDAP.
Specifically, the analyst can understand the process used by CEMDAP to generate the
activity-travel patterns by following flow represented by the Interactive UI windows. The
main dialog box of the interactive user interface looks as shown below.
Figure 4.23: Main Dialog Box of the Interactive User Interface
The Generation-Allocation Model System button opens up a dialog box within the
generation-allocation modeling framework as shown below with buttons for (1)
Generation and Timing of Work and School Activities of Adults and Children, (2)
Generation of Children’s Travel Needs and Allocation of Responsibilities to Parents, and
(3) Generation of Independent Activities by Adults and Children for Personal and
Household Serving Needs.
32
CEMDAP v 3.1 User’s Manual
Figure 4.24: Generation-Allocation Model System Interactive User Interface
Clicking on each of the buttons opens up the corresponding suite of individual
model components (shown in the following pages).
33
CEMDAP v 3.1 User’s Manual
Figure 4.25: Generation & Timing of Work and School Activities of Adults Interactive User
Interface
34
CEMDAP v 3.1 User’s Manual
Figure 4.26: Generation of Children’s Travel Needs and Allocation of Responsibilities to
Parents Interactive User Interface
35
CEMDAP v 3.1 User’s Manual
Figure 4.27: Generation of Independent Activities by Adults and Children for Personal and
Household Serving Needs Interactive User Interface
When all the components of the generation-allocation model system have been
configured, click Close to return to the main dialog box. The scheduling system user
interface (shown on the next page) can be obtained by clicking on the Scheduling Model
System button in the Overall Modeling Framework box. The Scheduling Model System
box has buttons for Work to Home Commute, Home to Work Commute, Non Worker Tour
Containing the Drop-off at School Activity, Non Worker Tour Containing the Pick-up at
School Activity, School-Home and Home-School Commutes, Worker’s Joint
Discretionary Tour with Children, Non Worker’s Joint Discretionary Tour with Children,
Independent Home-based and Work-based Tours for workers, Independent Home-based
Tours for non workers, and Independent Home-based Tours Undertaken by Children.
36
CEMDAP v 3.1 User’s Manual
Figure 4.28: Scheduling Model System Interactive User Interface
Clicking on each of the buttons opens up the corresponding suite of individual
model components (shown on the following pages). After configuring the individual
models of each suite, click the Close button to return to the Scheduling Model System
interface, in order to access the next suite of models.
37
CEMDAP v 3.1 User’s Manual
Figure 4.29: Scheduling of Work to Home Commute for Worker Interactive User Interface
38
CEMDAP v 3.1 User’s Manual
Figure 4.30: Scheduling of the Home to Work Commute for a Worker Interactive User Interface
39
CEMDAP v 3.1 User’s Manual
Figure 4.31: Scheduling of the Tour Containing the Drop-off Activity for a Nonworker
Interactive User Interface
40
CEMDAP v 3.1 User’s Manual
Figure 4.32: Scheduling of the Tour Containing the Pick-up Activity for a Nonworker
Interactive User Interface
Figure 4.33: Scheduling of the Tour Containing the Pick-up Activity for a Nonworker
Interactive User Interface
41
CEMDAP v 3.1 User’s Manual
Figure 4.34: Scheduling of the Joint Discretionary Tour Undertaken by a Worker with Children
Interactive User Interface
Figure 4.35: Scheduling of the Joint Discretionary Tour Undertaken by a Nonworker with
Children Interactive User Interface
42
CEMDAP v 3.1 User’s Manual
Figure 4.36: Scheduling of the Independent Home-based & Work-based Tours for Workers…
Interactive User Interface
Figure 4.37: Scheduling Independent Tours for Nonworkers Interactive User Interface
43
CEMDAP v 3.1 User’s Manual
Figure 4.38: Scheduling the Independent Discretionary Tour Undertaken by a Child Interactive
User Interface
When the entire scheduling model system for workers has been configured, return
to the main dialog box by clicking Close till the main dialog box appears.
4.4.3 Saving model configuration file
The model configurations entered using the dialog boxes can be saved into a model
configuration file for future use. This is achieved by clicking on the Save command in the
Models Menu.
44
CEMDAP v 3.1 User’s Manual
Figure 4.39: Save command in the Models Menu
The Save command opens the Save As dialog box. Choose the desired location and
type in the desired model configuration filename in this dialog box, and click Save to
finish.
Figure 4.40: Save As Dialog Box from the Save Command
45
CEMDAP v 3.1 User’s Manual
4.4.4 Loading model configuration file
In the current section instructions to load the prepared model configuration file are
provided. The model configuration file provided in the installation CD for testing
purposes (or saved using the Save command in the Models Menu, or created manually by
entering the model configurations into a text file), can be loaded into CEMDAP using the
Load command in the Models Menu.
Figure 4.41: Load Command in the Models Menu
The Load command opens the Save As dialog box (note that the model
configuration file must have already been saved as indicated in the previous section).
Navigate to the location of the model configuration file and select the file. Click Save to
finish loading the configuration file.
Note: For Testing the CEMDAP software, the reader is recommended to skip
to Section 4.5
46
CEMDAP v 3.1 User’s Manual
Figure 4.42: Save As Dialog Box from the Load Command
4.4.5 Configuring specific model types
All the model components in the embedded model system fall into one of seven
categories based on their econometric structures: Linear Regression, Hazard Duration,
Binary Logit, Multinomial Logit, Location Choice, Ordered Probit, or Work Start/End
Time (please see Guo et al. 2005 for a discussion on the structure of work start/end time
model, and Pinjari et al. 2006 for all the model components and their prescribed
econometric structures in CEMDAP). In this section, we take an example of each of these
model types and illustrate their methods of configuration. (Note: Clicking OK in any
dialog box only saves the information temporarily. In order to save a model
configuration, choose the Save command in the Models Menu).
Linear Regression
An example of a linear regression model is the Adult’s School Start Time. (Note
that the dependent variable in the regression models for duration is assumed to be the log
of the duration. This is done to ensure that we always predict a positive value).
The inputs required for this model are the variance parameter, the variable
parameters, and variables in the linear regression equation. First, enter the variance in the
box titled ‘Variance’ in the top left corner of the dialog box. Then enter the linear
regression equation by picking the relevant variable from the Variable list and entering
its parameter in the space designated ‘Coefficient’. Click the arrow button (=>) to add the
term to the linear regression expression. The ‘From’ option lists the sets of variables
available. ‘Households’ contains all the household-related variables from the input
47
CEMDAP v 3.1 User’s Manual
household table. ‘Persons’ contains the person-related variables from the input person
table and ‘Zones’ contains the zonal variables from the input zones table. ‘Other’ (which
is highlighted in the figure below) is a list of other variables including the constant, any
interaction terms and variables that may be generated during the simulation process by
models higher up in the modeling framework.
In the figure below, we enter the constant by selecting ‘Other’ in the ‘From’ list and
‘Constant (One)’ from the corresponding Variable list. Enter the value of the constant in
the Coefficient box and click the arrow button (=>) to enter the constant.
The reader will note that the number that appears at the beginning of variables in
the ‘Other’ list is the ID of the variable. It is the variable ID that gets stored in the model
specification file and it is the variable ID that is recognized by CEMDAP. (These IDs are
shown only for endogenous variables).
Figure 4.43: Linear Regression Model Interactive User Interface
When the linear regression equation has been completely configured, the dialog box
looks as shown below. The arrow buttons (=> and <=) can be used to add, remove, or edit
terms.
48
CEMDAP v 3.1 User’s Manual
Figure 4.44: Example of Fully Configured Linear Regression Model
Once the configuration of the individual models is finished, click OK to close the
model dialog box, and then click the Close button to return to the corresponding model
system interface, and then access the next suite of models. When the entire model system
has been configured, return to the main dialog box by clicking Close till the main dialog
box appears. This process is similar for all the individual models embedded in the model
system.
Hazard Duration
An example of a hazard duration model is the Children’s School Start Time model
within the generation-allocation model system. The model type supported is a hazardbased duration model with a non-parametric baseline and gamma heterogeneity. The
inputs required for this model can be configured in three steps. Step One: enter the
parameter estimated for the gamma heterogeneity term in the box titled ‘Variance’ in the
top left corner of the dialog box. Step Two: configure the parameters on the covariates.
This procedure is identical to entering the linear regression parameters. Select the
relevant input table from the ‘From’ list, and pick the variables from the Variable list.
Specify the coefficient for each variable and use the arrow button (=>) to add the terms.
49
CEMDAP v 3.1 User’s Manual
Figure 4.45: Hazard Duration Model Interactive User Interface
Step Three: after all the covariates have been configured (refer to the figure on the
next page), enter the threshold parameters. First, in the box titled ‘# Baseline time
intervals’ enter the number of discrete time periods (N) into which the duration is divided
(Note: Please do not enter zero because there cannot be zero number of discrete time
periods). Then, to enter the threshold parameters (number of threshold parameters = N-1),
click Edit. Additional boxes open up to enter the start times of each of the discrete
periods and the corresponding parameter estimates (sequentially from start time of the
second discrete period to the start time of the last discrete period, hence there are N-1
parameters). (Note: The end time of any discrete duration period is the start time of the
next period. The start time of the first discrete period is to be specified as zero, while the
end time of the last discrete period, which is not specified, is implicitly taken as infinity.).
Click on the arrow buttons (=> and <=) to scroll. When the hazard model is fully
configured, the dialog box looks as shown below. Note: If the user needs to change the
number of base line time intervals after clicking Edit, it is advised that the model
configuration be cancelled and reconfigured.
50
CEMDAP v 3.1 User’s Manual
Figure 4.46: Example of Fully Configured Hazard Duration Model
Multinomial Logit
An example of a multinomial logit model is the Commute Mode Choice Model
within the Worker’s Activity-Travel Scheduling model system. The discrete choices for
this model are drive alone, share auto ride as a passenger, share auto ride as a driver,
transit, and walk/bike. The model configuration dialog box for this looks as follows.
51
CEMDAP v 3.1 User’s Manual
Figure 4.47: Multinomial Logit Model Interactive User Interface
To configure the multinomial logit model, the utility functions of each of the
alternatives must be specified. Select the alternatives one at a time from the drop down
menu titled ‘Alternative’ to configure the corresponding utility function. In the example
shown above, we have selected the Drive alone alternative.
To configure the utility function for the drive alone alternative, select the relevant
variables one at a time and enter the coefficient in the designated space. Then click the
arrow button (=>) to add the term to the utility function for that alternative. This step is
similar to configuring the linear regression expression. In the example shown below we
have selected the constant and entered the corresponding coefficient for the utility
function of the drive alone alternative.
52
CEMDAP v 3.1 User’s Manual
Figure 4.48: Example of Constant and Corresponding Coefficient for the Utility Function of
one of the Multinomial Logit Model Alternatives
When the utility function for the alternative has been fully configured, the dialog
box looks as shown on the next page. The utility functions for all other alternatives can
be similarly configured. The utility of a base alternative with zero utility can be specified
by providing a value of zero for the constant of that alternative. After the utility functions
for all alternatives are configured, click on OK to complete the model configuration
specification. The reader should note that clicking on the Clear Function button clears all
the coefficients and variables so that the utility function can be reconfigured.
53
CEMDAP v 3.1 User’s Manual
Figure 4.49: Example of Fully Configured Multinomial Logit Model
Binary Logit
The binary logit model is a special case of the multinomial logit model, with only
two alternatives. It follows the same method of configuration as the multinomial logit.
The binary logit model, however, is simpler since it has only one utility function
(corresponding to the Yes alternative) to be configured. The No alternative is always
taken as the base with zero utility. For example, consider the Decision to go to Work
Model within the generation-allocation model system. The alternatives are either that the
person decides to go to work on the given day or not. Hence, there is only a single utility
function to be configured with the other alternative (No) serving as the base. The
completely configured dialog box looks as shown below. Note: Clicking on the Clear
Function button clears all the coefficients and variables so that the utility function and
can be reconfigured.
54
CEMDAP v 3.1 User’s Manual
Figure 4.50: Binary Logit Model Interactive User Interface
Location Choice
An example of a location choice model is the Work-Home Stop Location Model.
Since there is only one utility function to be configured, configuration of the location
choice model is similar to that of a linear regression model. First, enter the variance
determined for the travel time model corresponding to the tour under consideration in the
space designated ‘Variance’ (in this case enter the variance obtained for the travel time
model for stops in the Work-Home commute). Then configure the utility function in the
same way that the linear regression equation is configured. That is, select the relevant
input table from the ‘From’ list, and pick the variables from the ‘Variables’ box. Specify
the coefficient for each variable and use the arrow button (=>) to add the term. The
completely configured dialog box looks as shown below.
55
CEMDAP v 3.1 User’s Manual
Figure 4.51: Example of Fully Configured Binary Logit Model
Note that the list of variables supported by the location choice model is predetermined and available only under the ‘Other’ variables list. The list of variables
includes the inter-zonal LOS characteristics, and interactions between household or
person characteristics, and destination zone or inter-zonal LOS characteristics. The user
must use only these variables in the calibration and enter the parameters appropriately.
Ordered Probit
An example of an ordered probit model is the Work-Home Commute Number of
Stops Model. The inputs required for this model are the propensity function and threshold
parameters. Configuring the propensity function is identical to configuring the linear
regression equation. That is, select the relevant variables from each list (Households,
Persons, Zones and Other). Enter the coefficient corresponding to the variable and click
on the arrow button (=>) to add the term. The maximum number of alternatives to be
input for any model is pre-determined and based on the representation frameworks
supported for workers and non-workers. However, CEMDAP supports any number of
alternatives less than the maximum number. For example, the software supports a
maximum of four tours for non-workers in the D-FW model system. However, if the
maximum number of tours in a particular context is only 3, the number of threshold
parameters for the ordered probit model for # tours for non-workers is 2. This number
appears in the ‘# Threshold Values’ box in the bottom left corner of the dialog box. In the
Work-Home Commute Number of Stops Model, the software supports three stop options
56
CEMDAP v 3.1 User’s Manual
(0, 1, and 2 stops) and hence the number of threshold parameter is 2. Click on Edit to
specify the threshold parameters. The completely configured dialog box looks as shown
on the next page. Note: If the user needs to change the number of base line time intervals
after clicking Edit, it is advised that the model configuration be cancelled and
reconfigured.
Figure 4.52: Example of Fully Configured Ordered Probit Model
Work Start and End Time Model
The work start and end time model has a different model structure and
configuration than all other model structures discussed above (see Guo et al. 2005 for the
details on its model structure). The inputs required for this model (parameters and
variables) are entered just like the parameters and variables of the linear regression
model. Note that the list of variables supported by the work start and end time model is
pre-determined and available only under the ‘Other’ variables list. The user must use
only these variables in the calibration and enter the parameters appropriately.
57
CEMDAP v 3.1 User’s Manual
Figure 4.53: Work-Start-End-Time Model Interactive User Interface
4.5 INITIATING A SIMULATION RUN
Prior to starting a simulation, ensure that the input database is registered and
loaded, the output files are specified, and the model system is completely configured.
Then select the Run command in the Simulation Menu.
Figure 4.54: Run command in the Simulation Menu
58
CEMDAP v 3.1 User’s Manual
The progress bar is launched whenever a simulation run is initiated, and serves to keep
track of the progress of the simulation. In particular, the progress bar displays the number
of households processed, time elapsed, estimated simulation time left, number of threads
running, and the number of threads completed.
Figure 4.55: Simulation in Progress Dialog Box
The title of the progress bar reads ‘Simulation in Progress...’ when the simulation is
in progress and reads ‘Simulation Completed Successfully’ when it is completed.
Figure 4.56: Simulation Completed Successfully Dialog Box
Click OK and close the progress bar and ensure that the output files have been
created. The output files can then be imported into any spreadsheet and the simulated
activity-travel patterns of the population can be analyzed.
59
CEMDAP v 3.1 User’s Manual
60
CEMDAP v 3.1 User’s Manual
REFERENCES
Guo, J.Y., S. Srinivasan, N. Eluru, A. Pinjari, R. Copperman, and C.R. Bhat, "ActivityBased Travel-Demand Analysis for Metropolitan Areas in Texas: CEMSELTS
Model Estimations and Prediction Procedures, 4874 Zone System CEMDAP
Model Estimations and Procedures, and the SPG Software Details," Report 40807, prepared for the Texas Department of Transportation, October 2005.
A. Pinjari, N. Eluru, R. Copperman, I.N. Sener, J.Y. Guo, S. Srinivasan, and C.R. Bhat,
"Activity-Based Travel-Demand Analysis for Metropolitan Areas in
Texas: CEMDAP Models, Framework, Software Architecture and Application
Results, " Report 4080-8, prepared for the Texas Department
of Transportation, October 2006
61
CEMDAP v 3.1 User’s Manual
62
CEMDAP v 3.1 User’s Manual
Appendix A. D-FW Input Data
The input data requirements for running a simulation of the Dallas-Fort Worth (DFW) metropolitan area are dependent on the model configurations estimated for D-FW.
This input data therefore includes not only the ‘required variables’ specified in Chapter 3
but also additional variables that are specific to the estimated D-FW models. The folder
test_data, included in the /data directory under the installation destination directory,
contains the data required for a small sample of the D-FW population for test purposes.
The following tables contain the specifications for the input datasets.
Table A.1 Households Table Specification for D-FW Area
Column
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Variable Description
Household ID
Number of adults (age >=18)
Total number of HH vehicles, including motorcycles and RVs
Home TSZ location
Number of children
Household structure
Number of unemployed adults
Household income in 1000s of dollars
Household with no children
Number of persons in household
Multiple number of adults ( 1 if nadult > 1, 0 otherwise)
Vehicles per licensed driver
Single person household
Single parent household
Male-female couple household, no children
Male-female couple household, with children
Other household type
Number of licensed drivers
Household with one child
Household with two or more children
Number of employed adults
Household with no employed adult
Household with one employed adult
Household with two or more employed adults
Number of adult (age >= 18) students
Number of children (age < 18) who are not students
Number of children who are students
Own residential unit
Single family detached housing unit
Single family attached housing unit
Apartment type housing unit
Other type of housing unit
63
Variable
Name
HHID
NADULT
NVEH
HOMETSZ
NCHILD
HHSTRUCT
NAUNEMP
HHINCOME
ZEROCH
NPERS
MULTADU
VEHBYLIC
SPERSON
SPARENT
COUPLE
NUCLEAR
OHHTYPE
NUMLIC
ONECH
TWOCH
NAEMP
ZEROEMP
ONEEMP
TWOEMP
NASTU
NCNOTSTU
NCSTU
OWNHOME
SFDUNIT
SFAUNIT
APTUNIT
OTHUNIT
CEMDAP v 3.1 User’s Manual
Table A.2 Persons Table Specification for D-FW Area
Column
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Variable Description
Household ID
Person ID
Adult is employed
Adult or child is a student
Is the person licensed to drive
Work TSZ
School TSZ
Person is female
Age of the person in 1996
Adult is a parent
Caucasian
African American
Asian or pacific islander
Gender of the person
Personal vehicle availability (# vehicles/#Licensed Drivers)
High work flexibility
Pre school completed, child
Kindergarten to grade 4 completed, child
Grade 5 to grade 8 completed, child
Grade 9 or higher completed, child
Age of the person <= 5 years
Lowage * one employed adult in household
Lowage * two employed adults in household
Some college completed, adult
Associate or bachelors degree completed, adult
Masters or PhD degree completed
Income in 1000s of dollars
Income / HH income
Work duration between 0 and 20 hours a week
Work duration between 20 and 40 hours a week
Employment type: construction and manufacturing
Employment type: wholesale and transportation
Employment type: personal, professional and financial services
Employment type: public and military
Employment type: retail and repair
Female parent in a single parent or nuclear family household
Male parent in a single parent or nuclear family household
No school completed, child
Child not a student
Adult is unemployed
Adult son or daughter in a single parent or nuclear family household
Person is 16 years of age or older
Hispanic
Other race
Grade 8 or lower completed, adult
64
Variable
Name
HHID
PerID
Aemp
Stu
License
WorkTSZ
SchTSZ
Female
Age
Parent
Cauc
Afamer
Asian
Male
pvehavbl
Highflex
Presch
kgtog4
G5tog8
G9orhigh
Lowage
loage1
loage2
Somecol
Assobach
Mastphd
Income
Incomef
Wdurlow
Wdurmed
Emptype1
Emptype2
Emptype3
Emptype4
Emptype5
Mother
Father
Nosch
Cnotstu
Aunemp
Adchild
Adult
Hisp
Othrace
G8orlow
CEMDAP v 3.1 User’s Manual
46
47
48
49
50
51
52
Grade 9 to grade 12 completed, adult
High school completed, adult
Total weekly work duration (excluding weekend)
Work duration greater than 40 hours a week
Medium work flexibility
Low or no work flexibility
Employment type: other industries
G9tog12
Highsch
Workhrs
Wdurhigh
Medflex
Lowflex
Emptype6
Table A.3 Zones Table Specification for DF-W Area
Column
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Variable Description
TSZ zone id
Distance to the nearest major shopping zone
Accessibility with cost – retail employment
Accessibility to retail+ service employment
Accessibility to total employment
Accessibility to population
Dallas CBD dummy variable
Fort Worth CBD dummy variable
Median income of the zone (1000s of Dollars)
Number of households in the zone
Number of persons in the zone
Basic employment in the zone
Retail employment in the zone
Service employment in the zone
Total employment in the zone
Parking cost
County
Special land use
Internal zone dummy variable
Variable
Name
ZONEID
SHOPDIST
REMPACC
RSEMPACC
TEMPACC
POPACC
DALCBD
FWCBD
MEDINC
NUMHH
NUMPERS
BEMP
REMP
SEMP
TOTEMP
PARKCOST
COUNTY
SPLLUSE
INTERNAL
Note: The procedure to calculate the accessibility measures is provided at the end of this
appendix.
Table A.4 Zone2Zone Table Specification for DF-W Area
Column
#
1
2
3
4
Variable Description
TSZ zone id – origin
TSZ zone id – destination
Dummy variable indicating origin and destination are adjacent zones
Distance between zones
65
Variable
Name
Origin_zone
Dest_zone
Adjacent
Distance
CEMDAP v 3.1 User’s Manual
Table A.5 LOS Tables (AMPeak, PMPeak, and OFFPeak) Specifications for DF-W
Area
Column
#
1
2
3
4
5
6
7
8
9
10
11
12
Variable Description
TSZ zone id – origin
TSZ zone id – destination
Dummy variable indicating origin and destination are the same zone
Dummy variable indicating origin and destination are adjacent zones
Distance between zones
Auto In-vehicle travel time between origin and Destination zones
Auto Out-of-vehicle travel time between origin and Destination zones
Dummy variable indicating if transit is available between zones
Transit In-vehicle travel time between origin and Destination zones
Transit Out-of-vehicle travel time between origin and Destination zones
Transit cost
Auto cost
Variable
Name
Origin
Dest
samezone
Adjacent
Distance
autoIVTT
autoOVTT
Travail
TrIVTT
TrOVTT
Trcost
COST
Table A.6 LOSDir Tables (AMPeak, PMPeak, and OFFPeak) Specifications for DFW Area
Column
#
1
2
3
Variable
Name
start_t
End_t
tbname
Variable Description
Start Time of the time period
End Time of the time period
Name of the LOS table
Table A.7 LOSdir Table Currently Present in CEMDAP
Row#
1
2
3
4
5
start_t
0
360
930
210
780
End_t
210
780
1440
360
930
66
tbname
Losoffpk
Losoffpk
Losoffpk
lospeakam
lospeakpm
CEMDAP v 3.1 User’s Manual
Accessibility Measure Calculations
Accessibility to Retail Employment
Have to be re-evaluated when
1) retail employment changes, and/or
2) impedance measures change
Accessibility to retail employment measure for a time period t:
γ
⎛
⎞
1 N ⎜ ( retail_emp j ) ⎟
Acci ,t = ∑
N j =1 ⎜ ( Impedence ) β ⎟
ij,t
⎝
⎠
Impedenceij,t = IVTTij ,t + δ ⋅ Costij ,t
γ = 0.2868
δ = 0.0992
β = 3.0779
t : index for time period
i: index for origin zone
j: index for destination zone
IVTT in minutes
Cost in cents
N is the number of zones
The overall daily accessibility to retail employment measure is then computed as
1
∑ Durt ⋅ Acci,t
1440 t
Durt = Duration of time period t in minutes
Acci =
Use the following three TOD (time-of-day) periods: AM peak, off peak, PM peak
IF (TOD <= 209 | (TOD >=360 & TOD <= 719) | (TOD >= 930)) IVTT = off peak
IF (TOD >=210 & TOD <= 359) IVTT = AM peak.
IF (TOD >=720 & TOD <= 929) IVTT = PM peak.
TOD in minutes from 3 AM
Other accessibility measures
•
General formula = 1/N*SUM(size/impedance)
67
CEMDAP v 3.1 User’s Manual
•
Impedance = peak auto IVTT
•
Create 3 different accessibility measures using the following as the size measures:
(1) retail + service employment, (2) total employment, and (3) population
•
Sum over all internal zones, excluding the airport zones (if we don’t have the
internal identifier, sum over all zones except the airport zones). The airport zones
are identified in the special land use variable.
There measures have to be re-evaluated when the corresponding size measure changes.
68
CEMDAP v 3.1 User’s Manual
Appendix B. Input Database Preparation in PostgreSQL
Format
This document provides the instructions for setting up PostgreSQL and the
CEMDAP dataset on your computer.
1) Installing PostgreSQL
1. Go to the PostgreSQL on-line download page at:
http://www.postgresql.org/ftp/binary, and click on the latest folder of PostgreSQL
(i.e. the click on the folder that is farthest down the page that is not a beta version.
Note these instructions assume a version 8.2.4 is downloaded.). Next, locate the
win32 folder and click on it.
2. Next, locate and download the file postgresql-8.2.4-1.zip from one of the mirror
sites onto a temporary directory (e.g. C:\Temp) on your machine (either http or
ftp). (Note that the filename of the file changes with the version number).
3. Locate and extract the zip file.
4. Run the Windows Installer postgresql-8.2.msi (the file name varies with the
version number).
5. Follow the installation instructions listed on:
http://pginstaller.projects.postgresql.org/
Tips:
a) In the Service configuration page
1. Install as a service
2. Allow PostgreSQL to auto-generate the password. Hit NEXT by leaving
the Account password and Verify password section blank (the password
is not required for running PostgreSQL).
b) In the Initialize database cluster page
1. Leave the Initialize database cluster check box checked.
2. Choose the Superuser name (We recommend using postgres as the user
name)
3. Enter the password (we recommend using postgres as the password)
4. Please make sure you specify and write down your account name (say,
postgres) – this will be used as the User name – and the password. This
information will be required for accessing your database later.
c) In the Enable procedural languages page
1. Please accept the default settings and Hit Next
d) In the Enable contrib modules
69
CEMDAP v 3.1 User’s Manual
1. Please accept the default settings and Hit Next
e) In the PostGIS page
1. Please accept the default settings and Hit Next
2) Setting up the CEMDAP database
Once PostgreSQL is properly installed on your machine, follow
Start/Programs/PostgreSQL/pgAdmin III to run PostgreSQL. This opens up a database
maintenance window. You should be able to see in the database explorer (left-hand side
panel) the name of the database server (e.g. PostgreSQL Database Server 8.2) that you
created. To access the database, double click on the database server name. This will
prompt a dialog box for the password. Enter the user password you set up during the
installation process (Tip: have the password stored to avoid having to type it in every
time you want to access the database). After you enter the password, the Databases icon
will be visible under the database server name. To create a new database, right-click on
the Databases icon, and choose New Database. This will prompt a new dialog box. In this
dialog box, enter the name of the database and note down the given name. Note: The
name can be whatever the user desires (Ex: CEMDAP_Test_Data). Select Encoding as
“SQL_ASCII”, and leave the other fields empty. The procedure of loading data into this
empty database is described in the subsequent sections.
2.1. Create Tables
To access the empty database, click on the “+” symbol to the left of the database
name (Note you may have to double click on the database name the first time you access
the database. Press OK if a new window pops up). Then, you should be able to see items
such as Casts, Languages, Schemas, and Replication listed under the database name in
the Explorer. Expand Schemas and expand Public. You will see that the database
currently contains no tables (as indicated by Tables (0)). Therefore, the next step is to
create the skeleton of CEMDAP input tables. To do so:
1. Select Tools->Query Tool from the menu.
2. Paste the query statement provided below into the Query window and press F5 (or
select Query-Execute from the menu) to run the query. A message will appear
“Query returned successfully with no result…”.
3. Exit out of the Query window. Click NO to save changes.
The query creates empty tables (as can be seen under Tables) with the appropriate
column properties. These columns are empty, and so our next step is to populate the
columns.
70
CEMDAP v 3.1 User’s Manual
Query Statement for Table Creation
CREATE TABLE households
(
hid float8,
n_adults float8,
n_autos float8,
zone_id float8,
kids float8,
structure float8 DEFAULT 0,
naunemp float8,
hhincome float8,
zeroch float8,
npers float8,
multadu float8,
vehbylic float8,
sperson float8,
sparent float8,
couple float8,
nuclear float8,
ohhtype float8,
numlic float8,
onech float8,
twoch float8,
naemp float8,
zeroemp float8,
oneemp float8,
twoemp float8,
nastu float8,
ncnotstu float8,
ncstu float8,
ownhome float8,
sfdunit float8,
sfaunit float8,
aptunit float8,
othunit float8
)
WITH OIDS;
ALTER TABLE households OWNER TO postgres;
CREATE TABLE persons
(
hid float8,
pid float8,
employed float8,
71
CEMDAP v 3.1 User’s Manual
studying float8,
license float8,
work_zon float8,
stud_zon float8,
female float8,
age float8,
parent float8 DEFAULT 0,
caucasia float8,
afriamer float8,
asian float8,
male float8,
pvehavbl float8,
highflex float8,
presch float8,
kgtog4 float8,
g5tog8 float8,
g9orhigh float8,
lowage float8,
loage1 float8,
loage2 float8,
somecol float8,
assobach float8,
mastphd float8,
income float8,
incomef float8,
wdurlow float8,
wdurmed float8,
emptype1 float8,
emptype2 float8,
emptype3 float8,
emptype4 float8,
emptype5 float8,
mother float8,
father float8,
nosch float8,
cnotstu float8,
aunemp float8,
adchild float8,
adult float8,
hisp float8,
othrace float8,
g8orlow float8,
g9tog12 float8,
highsch float8,
workhrs float8,
wdurhigh float8,
72
CEMDAP v 3.1 User’s Manual
medflex float8,
lowflex float8,
emptype6 float8
)
WITH OIDS;
ALTER TABLE persons OWNER TO postgres;
CREATE TABLE losdir
(
start_t float8,
end_t float8,
tbname varchar(50)
)
WITH OIDS;
ALTER TABLE losdir OWNER TO postgres;
-- Table: losoffpk
-- DROP TABLE losoffpk;
CREATE TABLE losoffpk
(
orig_zon float8,
dest_zon float8,
same_zon float8,
adjacent float8,
distance float8,
da_ivtt float8,
da_ovtt float8,
tr_avail float8,
tr_ivtt float8,
tr_ovtt float8,
tr_cost float8,
au_cost float8
)
WITH OIDS;
-- Table: lospeakam
-- DROP TABLE lospeakam;
CREATE TABLE lospeakam
(
orig_zon float8,
dest_zon float8,
same_zon float8,
adjacent float8,
distance float8,
73
CEMDAP v 3.1 User’s Manual
da_ivtt float8,
da_ovtt float8,
tr_avail float8,
tr_ivtt float8,
tr_ovtt float8,
tr_cost float8,
au_cost float8
)
WITH OIDS;
ALTER TABLE lospeakam OWNER TO postgres;
ALTER TABLE losoffpk OWNER TO postgres;
-- Table: lospeakpm
-- DROP TABLE lospeakpm;
CREATE TABLE lospeakpm
(
orig_zon float8,
dest_zon float8,
same_zon float8,
adjacent float8,
distance float8,
da_ivtt float8,
da_ovtt float8,
tr_avail float8,
tr_ivtt float8,
tr_ovtt float8,
tr_cost float8,
au_cost float8
)
WITH OIDS;
ALTER TABLE lospeakpm OWNER TO postgres;
-- Table: zone2zone
-- DROP TABLE zone2zone;
CREATE TABLE zone2zone
(
orig_zon float8,
dest_zon float8,
adjacent float8,
distance float8
)
WITH OIDS;
ALTER TABLE zone2zone OWNER TO postgres;
74
CEMDAP v 3.1 User’s Manual
-- Table: zones
-- DROP TABLE zones;
CREATE TABLE zones
(
zid float8,
shopdist float8,
rempacc float8,
rsempacc float8,
tempacc float8,
popacc float8,
dalcbd float8,
fwcbd float8,
medinc float8,
numhh float8,
numpers float8,
bemp float8,
remp float8,
semp float8,
totemp float8,
parkcost float8,
county float8,
splluse float8,
internal float8
)
WITH OIDS;
ALTER TABLE zones OWNER TO postgres;
75
CEMDAP v 3.1 User’s Manual
2.2. Load Tables
The process of loading data files into the tables is done in DOS environment under
command prompt. To do so:
1. Open the Command Prompt window. To do this,
a. Click on Start/Run in windows
Figure B-1 Run Command in Start menu of Windows
b. Type cmd and click OK
Figure B-2 Run dialog Box
76
CEMDAP v 3.1 User’s Manual
2. At the command prompt, go to the \bin subdirectory under the Postgresql program
directory. To do this please follow the steps below:
a. At the command prompt type cd \ and press “Enter”.
Figure B-3 Command Prompt dialog box: cd\
b. Type cd Program files\postgresql\8.2\bin and press “Enter”. (Please note
that the version number 8.2 depends on the PostgreSQL version)
Figure B-4 Command Prompt dialog box: cd Program files\postgresql\8.2\bin
77
CEMDAP v 3.1 User’s Manual
3. Type in the following command, and press ‘Enter’:
psql –d database name –U user name
where database name is the name of the data base for which you are populating
data and user name is the user name you set up for accessing the database. For
example, if the database name is CEMDAP_Test_Data and user name is
postgres, the command should be:
psql –d CEMDAP_Test_Data –U postgres
Figure B-5 Command Prompt dialog box: psql –d database name –U user name
This command will direct you to the corresponding database with the database name as
defined above (the prompt is updated as the database name, in the case of our example it
becomes CEMDAP_Test_Data= #)
4. Once at the new prompt, type in the following command, and press ‘Enter’:
\copy name of the table from filename
For example, if the name of the table is households and the filename is
households.dat, the command should be as in the following:
78
CEMDAP v 3.1 User’s Manual
\copy households from ‘C:\\Program
Files\\PostgreSQL\\8.2\\bin\\TestData\\households.dat’
For testing CEMDAP, please copy the TestData folder from the installation CD
and place it in the C:\Program Files\PostgreSQL\8.2\bin. Tip: C:\\Program
Files\\PostgreSQL\\8.2\\bin\\TestData\\households.dat shows the location where
the data is loaded. In the example above, your data is located in the TestData
folder under C:\\Program Files\\PostgreSQL\\8.2\\bin. It is best to write out the
command line instead of copying and pasting.
Figure B-6 Command Prompt dialog box: C:\\Program
Files\\PostgreSQL\\8.2\\bin\\TestData\\households.dat
Tip: Note that name of the table corresponds to an empty table created in
pgAdmin III under the Table icon, and filename is the complete path name of the
file from which data is to be loaded.
5. Step four must be repeated for each table that needs to be populated with data (i.e.
losdir, losoffpk, lospeakam, lospeakpm, persons, zone2zone, and zones).
6. Exit out of the Command Prompt window.
2.3. Create Index
Once all files are loaded into the database, you will need to create indices for all
tables (households, losoffpk, etc.). To do so, you can follow two ways.
79
CEMDAP v 3.1 User’s Manual
One of the ways of creating indices can be performed by directly using a sql
statement to create the indexes for all the tables that you have. To do so:
1. In the pgADMIN window, Select Tools->Query Tool from the menu.
2. Paste the query statement provided below into the Query window and press F5 (or
select Query-Execute from the menu) to run the query. Once you run the query ( a
message will appear “Query returned successfully with no result…”), exit the Query
window (Click NO to saving the text), right click on the ‘Tables’ icon, and then click
“Refresh’ to see the indices updated.
CREATE UNIQUE INDEX households_1x
ON households
USING btree
(hid);
CREATE UNIQUE INDEX losdir_1x
ON losdir
USING btree
(start_t, end_t);
CREATE UNIQUE INDEX losoffpk_1x
ON losoffpk
USING btree
(orig_zon, dest_zon);
CREATE UNIQUE INDEX lospeakam_1x
ON lospeakam
USING btree
(orig_zon, dest_zon);
CREATE UNIQUE INDEX lospeakpm_1x
ON lospeakpm
USING btree
(orig_zon, dest_zon);
CREATE UNIQUE INDEX persons_1x
ON persons
USING btree
(pid, hid);
CREATE UNIQUE INDEX zone2zone_1x
ON zone2zone
USING btree
(orig_zon, dest_zon);
80
CEMDAP v 3.1 User’s Manual
CREATE UNIQUE INDEX zones_1x
ON zones
USING btree
(zid);
The second way of creating indexes is as follows:
1. Go back to the Database Explorer in pgAdmin III
2. Go to Schemas->Public->Tables
3. For each table listed under Tables:
a. Right click on ‘Indexes(0)’.
b. Select ‘New Index’ to open the New Index dialog box.
c. On the Properties tab page, enter the name of the index as “[name of the
file]_1x” (for example, enter “households_1x” for the households table).
Check Unique.
d. Next, go to the Columns page and add the column name(s) as shown in the
table below:
Table B.1 Column names of Postgres Tables
Table Name
Column(s) used for indexing
households
hid
Losdir
start_t, end_t
Losoffpk
orig_zon, dest_zon
Lospeakam
orig_zon, dest_zon
Lospeakpm
orig_zon, dest_zon
Persons
hid, pid
zone2zone
orig_zon,dest_zon
Zones
zid
If an index is set up correctly, you will see that Index (0) has become Index (1).
81
CEMDAP v 3.1 User’s Manual
82
CEMDAP v 3.1 User’s Manual
Appendix C. Structure of the Model Specification File
All the components of the model system embedded in CEMDAP must be
configured into a model configuration file (also referred to as a model specification file)
and input to the system before starting a simulation run. The preparation of model
configuration file involves providing the relevant variables and their parameters in a
specific format as explained below. Alternatively, a complete model specification file,
which can be used for testing purposes, is provided in the installation CD.
The specification file contains a complete description of each of the models
included in the CEMDAP modeling framework. The specification file starts with an
opening “<Configuration>” command and ends with a closing “</Configuration>”
command. Then, the individual models that are embedded in CEMDAP are coded
according to their specifications.
The first line of each model starts with the name of the model. Then, the class of
the model that identifies the modeling template from which the particular choice model is
derived (classCBLogitMM for Binary Logit, classCMNLogitMM for Multinomial Logit,
classCRegressMM for Linear Regression, classCOrdProbitMM for Ordered Probit,
classCHazardMM for Hazard Duration, classCBLogitMM for Location Choice, and
Work-Start-End-Time for Work Start and End Time Model), is recorded. The subsequent
lines then describe the IDs of each variable in the models and the parameter values
specified to those variables (a complete list of endogenous variables - variable IDs, and
explanations- used in the models can be found in the corresponding models in the GUI ).
To close the model, each opening statement should be closed (The reader will note that
each statement should start with an opening command and end with a closing command.
That is, the name and the class of the models are captured between the opening “<
ModelConfig>” command and the closing “</ModelConfig>” command. All the
variables in the model are captured between the opening “<Variables>” command and
the closing “</Variables>” command. If variance should be given in the model, the value
of variance is captured between the opening <Variance>” command and the closing
“</Variance>” command. Similarly, the alternatives in the Multinomial Logit model are
captured between the opening “<Alternatives>” command and closing “</Alternatives>”
command. The threshold values in the Ordered Probit model are captured between the
opening “<Thresholds>” command and closing “</Thresholds>” command. The interval
parameters for the Hazard Duration model are captured between the opening
“<Intervals>” command and closing “</Intervals>” command).
The following template represents the main structure of the specification file;
however, the reader will note that this template can be modified according to the different
specifications of each model embedded in the software. The number of variables can be
different for each model or some of the models can have different configurations.
83
CEMDAP v 3.1 User’s Manual
<Configuration>
<ModelConfig id="enter name of the model" optional-type="enter type of the model
*">
<Variables>
<Variable ID="enter the variable ID" value="enter the value of the corresponding
variable" optional-description="enter the description of the variable” />
</ Variables >
</ModelConfig>
< ModelConfig id ="enter name of the model" optional-type="enter type of the
model *">
< Variance>enter the value of the variance of the model</Variance>
<Variables>
<Variable ID="enter the variable ID" value="enter the value of the corresponding
variable" optional-description="enter the description of the variable” />
<Variable ID="enter the variable ID" value="enter the value of the corresponding
variable" optional-description="enter the description of the variable” />
</Variables>
</ModelConfig>
</Configuration>
In CEMDAP, there exist seven different choice models: 1) Binary Logit, 2)
Multinomial Logit, 3) Linear Regression, 4) Ordered Probit, 5) Hazard Duration, 6)
Location Choice, and 7) Work Start and End Time. The format of the entry in the model
specification file, for each of the seven models is slightly different. Below, the
specification file format for each of the seven models is given in detail.
1) Binary Logit Model
The first line in the Binary Logit Model starts with the name of the model that is coded in
the ModelConfig id part. The class of the model (classCBLogitMM) is also identified in
the first line of the configuration in the optional-type part. The second line in the syntax
indicates an opening command for the variables of the model. The subsequent lines then
describe the various parameters associated with that model. In order to describe the
variables, the Variable ID of each variable is first entered in the Variable ID part, and
then the corresponding parameter of the variable is coded in the value part. It is also
possible to give a description for the variable in the optional-description part. Once the
entire set of variables embedded in the model are defined, the model is closed (The reader
will note that each row under the “<Variables>” command represents a single variable.)
The following configuration presents the framework for one of the Binary Logit models
in CEMDAP: Children’s decision to go to school. The first variable for the “Children’s
decision to go to school” model is constant that has a variable ID of 30000 and the
corresponding value of the parameter is -0.5765.
84
CEMDAP v 3.1 User’s Manual
<ModelConfig id="Children's decision to go to school" optional-type="class
CBLogitMM *">
<Variables>
<Variable ID="30000" value="-0.5765" optional-description="unavailable" />
<Variable ID="10016" value="0.9046" optional-description="unavailable" />
<Variable ID="10017" value="1.9349" optional-description="unavailable" />
<Variable ID="10018" value="1.8628" optional-description="unavailable" />
<Variable ID="10019" value="1.6204" optional-description="unavailable" />
<Variable ID="7" value="0.0056" optional-description="unavailable" />
</Variables>
</ModelConfig>
2) Multinomial Logit Model
The first line in the Multionomial Logit Model starts with the name of the model that is
coded in the ModelConfig id part. The class of the model (classCMNLogitMM) is also
identified in the first line of the configuration in the optional-type part. The second line in
the syntax is the ID of the base alternative (which should be always zero). The
subsequent lines then describe the various parameters associated with Multinomial Logit
model for each alternative. In order to describe the variables in each alternative define the
ID of the alternative in the Alternative ID part. Then, the variables are described in the
Variables part under the corresponding alternative. The Variable ID of each variable is
first entered in the Variable ID part, and then the corresponding parameter of the variable
is coded in the value part. It is also possible to give a description for the variable in the
optional-description part. Once the entire set of variables for each alternative in the
model are defined, the model is closed (The reader will note that each row under the
“<Variables>” command represents a single variable for the corresponding alternative.)
The following configuration presents the framework for one of the Multinomial Logit
models in CEMDAP: Allocation of drop off episode to parent. The model has 3
alternatives (Alternative 0 (base), Alternative 1, and Alternative 2). The first variable in
Alternative 1 is constant that has a variable ID of 30000 and the corresponding value of
the parameter is -0.5807
<ModelConfig id="Allocation of drop off episode to parent" optional-type="class
CMNLogitMM *">
<Alternatives baseID="0">
<Alternative ID="0">
<Variables />
</Alternative>
<Alternative ID="1">
<Variables>
<Variable ID="30000" value="-0.5807" optional-description="unavailable" />
<Variable ID="30013" value="0.0041" optional-description="unavailable" />
<Variable ID="30015" value="-0.0047" optional-description="unavailable" />
<Variable ID="10032" value="0.9955" optional-description="unavailable" />
</Variables>
85
CEMDAP v 3.1 User’s Manual
</Alternative>
<Alternative ID="2">
<Variables>
<Variable ID="30013" value="0.0041" optional-description="unavailable" />
<Variable ID="30015" value="-0.0047" optional-description="unavailable" />
<Variable ID="10032" value="0.9955" optional-description="unavailable" />
</Variables>
</Alternative>
</Alternatives>
</ModelConfig>
3) Linear Regression Model
The first line in the Linear Regression Model starts with the name of the model that is
coded in the ModelConfig id part. The class of the model (classCRegressMM) is also
identified in the first line of the configuration in the optional-type part. The second line in
the syntax specifies the variance value of the model (if the value of the variance is zero,
then the user should enter a very small value for variance such as 0.00001). The third line
indicates an opening command for the variables of the model. The subsequent lines then
describe the various parameters associated with that model. In order to describe the
variables, the Variable ID of each variable is first entered in the Variable ID part, and
then the corresponding parameter of the variable is coded in the value part. It is also
possible to give a description for the variable in the optional-description part. Once the
entire set of variables embedded in the model are defined, the model is closed (The reader
will note that each row under the “<Variables>” command represents a single variable.)
The following configuration presents the framework for one of the Linear Regression
models in CEMDAP: Adult’s school start time (time from 3 AM). The variance of the
model has a value of 0.0764. The first variable of the model is constant that has a
variable ID of 30000 and the corresponding value of the parameter is 5.7896.
<ModelConfig id="Adult's school start time (time from 3 AM)" optionaltype="class CRegressMM *">
<Variance>0.0764</Variance>
<Variables>
<Variable ID="30000" value="5.7896" optional-description="temp" />
<Variable ID="10023" value="0.1696" optional-description="temp" />
<Variable ID="10024" value="0.1696" optional-description="temp" />
<Variable ID="10025" value="0.2757" optional-description="temp" />
<Variable ID="10040" value="-0.1389" optional-description="temp" />
<Variable ID="10026" value="0.0011" optional-description="temp" />
<Variable ID="16" value="-0.1280" optional-description="temp" />
</Variables>
</ModelConfig>
4) Ordered Probit Model
The first line in the Ordered Probit Model starts with the name of the model that is coded
in the ModelConfig id part. The class of the model (classCOrdProbitMM) is also
86
CEMDAP v 3.1 User’s Manual
identified in the first line of the configuration in the optional-type part. The second line
indicates an opening command for the variables of the model. The subsequent lines then
describe the various parameters associated with that model. In order to describe the
variables the Variable ID of each variable is first entered in the Variable ID part, and then
the corresponding parameter of the variable is coded in the value part. It is also possible
to give a description for the variable in the optional-description part. Once the entire set
of variables embedded in the model are defined, variables part is closed (The reader will
note that each row under the “<Variables>” command represents a single variable.). The
following line indicates an opening command for the threshold values of the model. The
values of each threshold are entered in the Threshold value part one by one. Once the
entire threshold values are entered, the model is closed. The following configuration
presents the framework for one of the Ordered Probit models in CEMDAP: Number of
stops in a tour (before-work tour). The first variable of the model has a variable ID of
30085 and the corresponding value of the parameter is –0.728. There are four threshold
parameters in this model; and their values are 3.1768, 4.0896, 4.7494 and 5.194.
<ModelConfig id="Number of Stops in a tour (before-work tour)" optionaltype="class COrdProbitMM *">
<Variables>
<Variable ID="30085" value="-0.728" optional-description="unavailable" />
<Variable ID="30104" value="0.402" optional-description="unavailable" />
<Variable ID="30092" value="0.0024" optional-description="unavailable" />
<Variable ID="30093" value="0.763" optional-description="unavailable" />
<Variable ID="30094" value="0.7391" optional-description="unavailable" />
<Variable ID="30096" value="0.5848" optional-description="unavailable" />
<Variable ID="30070" value="-0.2946" optional-description="unavailable" />
<Variable ID="30072" value="-0.3462" optional-description="unavailable" />
<Variable ID="30029" value="0.8791" optional-description="unavailable" />
<Variable ID="30041" value="0.6869" optional-description="unavailable" />
<Variable ID="30043" value="1.0051" optional-description="unavailable" />
<Variable ID="30044" value="0.6032" optional-description="unavailable" />
<Variable ID="30045" value="0.5741" optional-description="unavailable" />
<Variable ID="30046" value="0.8444" optional-description="unavailable" />
<Variable ID="10002" value="-0.2135" optional-description="unavailable" />
<Variable ID="12" value="0.2483" optional-description="unavailable" />
<Variable ID="30022" value="0.1465" optional-description="unavailable" />
</Variables>
<Thresholds>
<Threshold value="3.1768" />
<Threshold value="4.0896" />
<Threshold value="4.7494" />
<Threshold value="5.194" />
</Thresholds>
</ModelConfig>
87
CEMDAP v 3.1 User’s Manual
5) Hazard Duration Model
The first line in the Hazard Duration Model starts with the name of the model that is
coded in the ModelConfig id part. The class of the model (classCHazardMM) is also
identified in the first line of the configuration in the optional-type part. The second line in
the syntax specifies the variance value of the model. The third line indicates an opening
command for the variables of the model. The subsequent lines then describe the various
parameters associated with that model. In order to describe the variables the Variable ID
of each variable is first entered in the Variable ID part, and then the corresponding
parameter of the variable is coded in the value part. It is also possible to give a
description for the variable in the optional-description part. Once the entire set of
variables embedded in the model are defined, variables section is closed (The reader will
note that each row under the “<Variables>” command represents a single variable). The
following line indicates an opening command for the intervals that are specified in the
model. The interval start time and corresponding parameter is entered for each of the
interval in the model. Once all the interval characteristics are defined, the model is
closed. The following configuration presents the framework for one of the Hazard
Duration models in CEMDAP: Children’s school start time (time from 3 AM). The
variance of the model has a value of 0.0310. The first variable of the model has a variable
ID of 10020 and the corresponding value of the parameter is 0.5034. The first interval
time for the model is 0, and the corresponding parameter is -2.5892.
<ModelConfig id="Children's school start time (time from 3 AM)" optionaltype="class CHazardMM *">
<Variance>0.0310</Variance>
<Variables>
<Variable ID="10020" value="0.5034" optional-description="unavailable" />
<Variable ID="10017" value="-0.2604" optional-description="unavailable" />
<Variable ID="10011" value="-0.2393" optional-description="unavailable" />
<Variable ID="10012" value="0.8228" optional-description="unavailable" />
<Variable ID="6" value="0.1314" optional-description="unavailable" />
</Variables>
<Intervals>
<Interval start="0" param="-2.5892" />
<Interval start="260.5" param="-1.9999" />
<Interval start="270.5" param="-1.4535" />
<Interval start="280.5" param="-0.9721" />
<Interval start="285.5" param="-0.6452" />
<Interval start="290.5" param="-0.4148" />
<Interval start="295.5" param="-0.0264" />
<Interval start="300.5" param="0.2779" />
<Interval start="310.5" param="0.5515" />
<Interval start="320.5" param="0.7849" />
<Interval start="330.5" param="1.0679" />
<Interval start="350.5" param="1.3303" />
</Intervals>
</ModelConfig>
88
CEMDAP v 3.1 User’s Manual
6) Location Choice Model
The first line in the Location Choice Model starts with the name of the model that is
coded in the ModelConfig id part. The class of the model (classCLocationMM) is also
identified in the first line of the configuration in the optional-type part. The second line in
the syntax indicates an opening command for the variables of the model. The subsequent
lines then describe the various parameters associated with that model. In order to describe
the variables the Variable ID of each variable is first entered in the Variable ID part, and
then the corresponding parameter of the variable is coded in the value part. It is also
possible to give a description for the variable in the optional-description part. Once the
entire variables embedded in the model are defined, the model is closed (The reader will
note that each row under the “<Variables>” command represents a single variable). The
following configuration presents the framework for one of the Location Choice models in
CEMDAP: Location choice: stops in a workers tour/commute. The first variable of the
model has a variable ID of 30165 and the corresponding value of the parameter is
-0.2495.
<ModelConfig id="location choice: stops in a workers tour/commute" optionaltype="classCLocationMM *">
<Variables>
<Variable ID="30165" value="-0.2495" optional-description="unavailable" />
<Variable ID="30166" value="-0.6848" optional-description="unavailable" />
<Variable ID="30167" value="-0.1684" optional-description="unavailable" />
<Variable ID="30168" value="-0.163" optional-description="unavailable" />
<Variable ID="30169" value="0.4021" optional-description="unavailable" />
<Variable ID="30170" value="1.2076" optional-description="unavailable" />
<Variable ID="30171" value="-1.259" optional-description="unavailable" />
<Variable ID="30172" value="0.2544" optional-description="unavailable" />
<Variable ID="30173" value="0.2024" optional-description="unavailable" />
<Variable ID="30174" value="0.1577" optional-description="unavailable" />
<Variable ID="30175" value="0.2264" optional-description="unavailable" />
<Variable ID="30176" value="0.2287" optional-description="unavailable" />
</Variables>
</ModelConfig>
7) Work Start and End Time Model
The first line in the Work Start and End Time Model starts with the name of the model
that is coded in the ModelConfig id part. The class of the model (class
CStartAndEndTimeMM) is also identified in the first line of the configuration in the
optional-type part. The second line in the syntax indicates an opening command for the
variables of the model. The subsequent lines then describe the various parameters
associated with that model. In order to describe the variables the Variable ID of each
variable is first entered in the Variable ID part, and then the corresponding parameter of
the variable is coded in the value part. It is also possible to give a description for the
variable in the optional-description part. Once the entire set of variables embedded in the
89
CEMDAP v 3.1 User’s Manual
model are defined, the model is closed (The reader will note that each row under the
“<Variables>” command represents a single variable.) The following configuration
presents the framework for the Work Start and End Time model in CEMDAP. The first
variable of the model has a variable ID of 30264 and the corresponding value of the
parameter is -1.8751.
<ModelConfig id="Work-Start-End-Time" optional-type="class
CStartAndEndTimeMM *">
<Variables>
<Variable ID="30264" value="-1.8751" optional-description="unavailable" />
<Variable ID="30265" value="2.3648" optional-description="unavailable" />
<Variable ID="30266" value="1.0647" optional-description="unavailable" />
<Variable ID="30267" value="-7.9309" optional-description="unavailable" />
<Variable ID="30268" value="-4.499" optional-description="unavailable" />
<Variable ID="30269" value="-1.4412" optional-description="unavailable" />
<Variable ID="30270" value="6.9739" optional-description="unavailable" />
<Variable ID="30271" value="3.6043" optional-description="unavailable" />
<Variable ID="30272" value="0.7273" optional-description="unavailable" />
<Variable ID="30273" value="-4.3365" optional-description="unavailable" />
<Variable ID="30274" value="-0.7668" optional-description="unavailable" />
<Variable ID="30275" value="0.1056" optional-description="unavailable" />
<Variable ID="30015" value="82.4853" optional-description="unavailable" />
<Variable ID="30018" value="-0.0295" optional-description="unavailable" />
</Variables>
</ModelConfig>
90
">