Working with Predictive Planning in Smart View

Oracle® Cloud
Working with Predictive Planning in Smart
View
E94164-01
Oracle Cloud Working with Predictive Planning in Smart View,
E94164-01
Copyright © 2015, 2018, Oracle and/or its affiliates. All rights reserved.
Primary Author: EPM Information Development Team
This software and related documentation are provided under a license agreement containing restrictions on
use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your
license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify,
license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means.
Reverse engineering, disassembly, or decompilation of this software, unless required by law for
interoperability, is prohibited.
The information contained herein is subject to change without notice and is not warranted to be error-free. If
you find any errors, please report them to us in writing.
If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on
behalf of the U.S. Government, then the following notice is applicable:
U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software,
any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users are
"commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agencyspecific supplemental regulations. As such, use, duplication, disclosure, modification, and adaptation of the
programs, including any operating system, integrated software, any programs installed on the hardware,
and/or documentation, shall be subject to license terms and license restrictions applicable to the programs.
No other rights are granted to the U.S. Government.
This software or hardware is developed for general use in a variety of information management applications.
It is not developed or intended for use in any inherently dangerous applications, including applications that
may create a risk of personal injury. If you use this software or hardware in dangerous applications, then you
shall be responsible to take all appropriate fail-safe, backup, redundancy, and other measures to ensure its
safe use. Oracle Corporation and its affiliates disclaim any liability for any damages caused by use of this
software or hardware in dangerous applications.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of
their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are
used under license and are trademarks or registered trademarks of SPARC International, Inc. AMD, Opteron,
the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced Micro
Devices. UNIX is a registered trademark of The Open Group.
This software or hardware and documentation may provide access to or information about content, products,
and services from third parties. Oracle Corporation and its affiliates are not responsible for and expressly
disclaim all warranties of any kind with respect to third-party content, products, and services unless otherwise
set forth in an applicable agreement between you and Oracle. Oracle Corporation and its affiliates will not be
responsible for any loss, costs, or damages incurred due to your access to or use of third-party content,
products, or services, except as set forth in an applicable agreement between you and Oracle.
Contents
Documentation Accessibility
Documentation Feedback
1
Getting Started
Overview
1-1
Installing Predictive Planning
1-1
Checking for Updates
1-2
Uninstalling Predictive Planning
1-2
Starting Predictive Planning
1-2
The Predictive Planning Ribbon
1-3
Running a Standard Prediction
1-3
Using Quick Predict
1-4
Quick Predict Example 1
1-5
Quick Predict Example 2
1-6
Predictive Planning for Users of Ad Hoc Grids
2
1-6
Viewing Results
Using the Predictive Planning Panel
2-1
Chart Tab
2-2
Data Tab
2-3
Statistics Tab
2-4
Summary Area
2-6
Setting Chart Preferences
2-6
Adjusting Future Data Series
2-7
Adjusting Future Series with the Mouse
2-7
Using the Adjust Series Dialog
2-9
Using Comparison Views
Editing the Current View
Adding a Scenario
2-10
2-11
2-12
iii
3
Adding Prediction Data
2-12
Adding a Trend Line
2-12
Creating a New View
2-13
Managing Views
2-13
Analyzing Results
Overview
3-1
Filtering Results
3-1
Pasting Results
3-2
Creating Reports
3-3
Setting Report Preferences
Extracting Data
Setting Data Extraction Preferences
4
Setting General Predictive Planning Options
A
Setting Up Predictive Planning
Before You Begin
3-4
3-4
3-5
A-1
Assigning Security Roles
A-1
Hierarchical Data Prediction Issues
A-1
Comparing Bottom-up, Top-down, and Full Predictions
A-1
Pasting Results for Predictions
A-2
Aggregating Best and Worst Case Predictions
A-2
Historical Data and Prediction Accuracy
A-3
Form Creation and Modification Issues
A-3
Using Valid Forms
A-3
Determining the Time Granularity of Predictions
A-3
Determining the Prediction Range
A-4
Creating a New Scenario for Prediction Results
A-4
Setting Form Defaults
A-5
Application and Individual Form Defaults
A-5
Using the Set Up Prediction Dialog
A-6
Specifying a Historical Data Source
A-6
Mapping Member Names
A-8
About Name Defaults
A-9
Selecting Members
A-9
Setting Prediction Options
A-10
Using Alternate Historical Data Sources
A-12
Alternate Plan Types and POV Configuration
A-13
iv
Alternate Plan Types and Dates
B
A-13
Predictive Planning Forecasting and Statistical Descriptions
Classic Time-series Forecasting
Classic Nonseasonal Forecasting Methods
B-1
B-1
Single Moving Average (SMA)
B-1
Double Moving Average (DMA)
B-2
Single Exponential Smoothing (SES)
B-2
Double Exponential Smoothing (DES)
B-3
Damped Trend Smoothing (DTS) Nonseasonal Method
B-3
Classic Nonseasonal Forecasting Method Parameters
B-3
Classic Seasonal Forecasting Methods
B-4
Seasonal Additive
B-4
Seasonal Multiplicative
B-4
Holt-Winters’ Additive
B-5
Holt-Winters’ Multiplicative
B-5
Damped Trend Additive Seasonal Method
B-6
Damped Trend Multiplicative Seasonal Method
B-6
Classic Seasonal Forecasting Method Parameters
B-7
ARIMA Time-series Forecasting Methods
B-7
Time-series Forecasting Error Measures
B-8
RMSE
Forecasting Method Selection
B-8
B-8
v
Documentation Accessibility
Documentation Accessibility
For information about Oracle's commitment to accessibility, visit the Oracle
Accessibility Program website at http://www.oracle.com/pls/topic/lookup?
ctx=acc&id=docacc.
Access to Oracle Support
Oracle customers that have purchased support have access to electronic support
through My Oracle Support. For information, visit http://www.oracle.com/pls/topic/
lookup?ctx=acc&id=info or visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs
if you are hearing impaired.
vi
Documentation Feedback
To provide feedback on this documentation, send email to epmdoc_ww@oracle.com,
or, in an Oracle Help Center topic, click the Feedback button located beneath the
Table of Contents (you may need to scroll down to see the button).
Follow EPM Information Development on these social media sites:
LinkedIn - http://www.linkedin.com/groups?gid=3127051&goback=.gmp_3127051
Twitter - http://twitter.com/hyperionepminfo
Facebook - http://www.facebook.com/pages/Hyperion-EPM-Info/102682103112642
Google+ - https://plus.google.com/106915048672979407731/
#106915048672979407731/posts
YouTube - https://www.youtube.com/oracleepminthecloud
vii
1
Getting Started
Related Topics
•
Overview
•
Installing Predictive Planning
•
Starting Predictive Planning
•
The Predictive Planning Ribbon
•
Running a Standard Prediction
•
Using Quick Predict
•
Predictive Planning for Users of Ad Hoc Grids
Overview
The Predictive Planning feature of Planning is an extension to Oracle Smart View for
Office that works with valid Planning forms to predict performance based on historical
data. Predictive Planning uses sophisticated time-series forecasting techniques to
create new predictions or validate existing forecasts that were entered into Planning
using other forecasting methods.
Notes about working with Predictive Planning:
•
Predictive Planning is currently available in 32-bit and 64-bit implementations.
•
Valid ad hoc grids are supported. For details, see Predictive Planning for Users of
Ad Hoc Grids.
•
Predictive Planning supports sandbox versions for forms and adhoc grids.
Historical data is read from the same sandbox that the form or adhoc grid is using.
•
Users with security roles that enable them to modify Planning forms should read
Setting Up Predictive Planning in this Guide to ensure that forms are configured
for maximum compatibility.
Watch this overview video to learn about using Predictive Planning with Oracle
Planning and Budgeting Cloud.
Overview Video
Installing Predictive Planning
To install Predictive Planning, follow the instructions in Using Oracle Planning and
Budgeting Cloud Service.
1-1
Chapter 1
Starting Predictive Planning
Checking for Updates
Access to recent features in Predictive Planning is dependent on having the latest
Oracle Smart View for Office release.
If your Administrator advises, update Predictive Planning by downloading and
installing the latest version of Predictive Planning using one of these methods:
•
In Microsoft Excel, choose Smart View, then Options, and then Extensions.
Click Check For Updates, and if the Predictive Planning extension displays
Update Available, click to download and then install the latest version. You will be
prompted to close all Microsoft Office applications.
•
Install the latest release of Predictive Planning from the Oracle Planning and
Budgeting Cloud Home page.
•
Download and install the latest version of Predictive Planning from the location
your administrator advises.
Tip:
To determine the version of Predictive Planning that you are using, from the
Predictive Planning ribbon, select Help, and then select About.
Uninstalling Predictive Planning
To uninstall Predictive Planning:
1.
If your Oracle Smart View for Office Administrator has enabled the uninstall option,
in Microsoft Excel, choose Smart View, then Options, then Extensions, and then
click Remove next to the Predictive Planning extension.
2.
If the Remove button is not available, then use Windows Add/Remove programs
(or Programs and Features) to uninstall.
Starting Predictive Planning
To start Predictive Planning:
1.
Confirm that compatible versions of Oracle Smart View for Office, Predictive
Planning, and Microsoft Excel are installed on your computer and that you have
access to a compatible version of Planning.
2.
Start Microsoft Excel.
3.
In Smart View, connect to a source.
4.
Open a valid Planning form (Using Valid Forms).
5.
To display the Predictive Planning ribbon, select the Planning ribbon, and then
click Predict.
1-2
Chapter 1
The Predictive Planning Ribbon
The Predictive Planning Ribbon
When you start Predictive Planning, the Predictive Planning ribbon is added to the
ribbon bar.
Figure 1-1
Predictive Planning Ribbon
Button groups are as follows:
•
Run—Sets form preferences and runs predictions
•
View—Displays and manages views of results
•
Analyze—Filters and pastes results, creates reports, and extracts data to the
spreadsheet
•
Help—Displays online help and information about this version of Predictive
Planning.
A tooltip identifies each button when you point to it.
For online help and information about Predictive Planning, select Help, and then
Predictive Planning.
For a list of shortcut keys (keyboard equivalents of buttons and commands), see the
Accessibility Guide for Oracle Planning and Budgeting Cloud Service.
You can use Predictive Planning in two ways:
•
Running a Standard Prediction
•
Using Quick Predict
Running a Standard Prediction
When you run a prediction, Predictive Planning analyzes historical data for each
selected member, and then projects this information into the future to generate
predicted results. If the Planning administrator has created a scenario for the predicted
data, you can paste it into Oracle Smart View for Office without overwriting existing
data.
To run a standard prediction:
1.
Select the Predictive Planning ribbon (The Predictive Planning Ribbon).
2.
Select Predict,
.
1-3
Chapter 1
Using Quick Predict
3.
Review the Run Confirmation dialog.
It shows the number of members, the source and range of historical data to
include in the prediction, and the predicted date range.
4.
Optional: View or change included members and the historical or predicted date
range.
•
By default, all editable members are selected. To change this, click Change
and see Selecting Members.
•
By default, predictions are based on all historical data for a series. To select a
specific data range for historical or predicted data, click Change and then
specify a start and end year and time period.
Note:
For the most accurate predictions, the number of periods of historical
data available should be at least twice the number of prediction periods
requested. If you have specified more prediction periods, you are
prompted to reduce the number.
5.
When the displayed settings are complete, click Run.
6.
Review the Run Summary dialog, if present, and click OK.
Results are displayed in the Predictive Planning panel. By default, the Chart tab is
selected (Figure 1).
Using Quick Predict
When you run a prediction, Predictive Planning analyzes historical data for each
selected member, and then projects this information into the future to generate
predicted results. With Quick Predict, all form defaults, except those for member
selection, are used without displaying dialogs. The predicted results are immediately
pasted into the Planning form. You can choose whether to enter predicted data into all
cells for a member or just selected cells.
Note:
To avoid overwriting existing data, the Planning administrator should add a
prediction scenario to the form before you predict data.
To run a prediction with Quick Predict:
1.
In a Planning form in Oracle Smart View for Office, select member names or cells
to predict.
2.
Right-click and then select Predictive Planning,
1-4
Chapter 1
Using Quick Predict
or select the Predictive Planning ribbon (The Predictive Planning Ribbon), and
then click the lower half of the Predict button,
3.
, with the label and arrow.
Indicate whether to predict an entire member or just selected cells:
•
Select Quick Predict Selected Members to predict future values for selected
members and paste results into all the members' future data cells.
•
Select Quick Predict To Selected Cells to predict future values for members
containing the selected data cells and paste results into only the selected
cells.
Note:
If the selection includes more than one scenario, you are prompted to
choose one for the target of the cell pasting.
Results are pasted as requested. Success icons and prediction quality values are
displayed for selected members in the column to the right of the last column of data
For examples, click the listed links.
Results are not displayed in the Smart View panel by default. To display a chart and
other results, open the list next to the Home icon in the Smart View panel, and then
select Predictive Planning. Initially, the Chart tab is selected (Figure 1). Whatever
results tab you viewed last is displayed.
Quick Predict Example 1
In Figure 1, the user selected cells in the Prediction row for two members for months
beyond the actual data. Then, the user selected Quick Predict To Selected Cells.
The predicted data was pasted into the selected cells.
Figure 1-2
Quick Predict Example 1, Pasting to Selected Cells
1-5
Chapter 1
Predictive Planning for Users of Ad Hoc Grids
Quick Predict Example 2
In Figure 1, the user selected three member names were selected, and then selected
Quick Predict Selected Members. Because the selection included multiple scenarioversion choices, the user had to respond to a prompt. Then, predicted values were
pasted into the Prediction version for the Boom Box and Personal CD Player
members.
Figure 1-3
Members
Quick Predict Example 2, Pasting Predicted Values for Selected
Predictive Planning for Users of Ad Hoc Grids
You can use Predictive Planning with ad hoc grids as well as standard Planning forms.
When you open a valid ad hoc grid with Predictive Planning installed, the Predict
button is displayed on the Planning Ad Hoc ribbon. Click it to display the Predictive
Planning ribbon (The Predictive Planning Ribbon). Controls work the same as for
standard Planning forms. You can use Quick Predict or run standard predictions
(Using Quick Predict). The special charting features also are available (Adjusting
Future Data Series).
All Predictive Planning functionality works with ad hoc grids, considering the following:
•
If you enter free-form mode, you must click Refresh before you run a prediction.
•
When you create an ad hoc grid, any Predictive Planning preferences that were
available in the original Planning form are applied to the new ad hoc grid. If you
create an ad hoc grid without starting from a Planning form, the default
preferences from the application are used.
•
You can set preferences freely through the Set Up Prediction button without
Administrator rights. However, you can save preferences only by saving the ad
hoc grid, if your security role allows that.
1-6
Chapter 1
Predictive Planning for Users of Ad Hoc Grids
•
Ad hoc grids have the same validation requirements as standard forms (Using
Valid Forms). If a form is not valid for Predictive Planning, Predict does not
display on the Planning Ad Hoc ribbon (unless the Show ribbon only for valid
Planning forms option is disabled).
1-7
2
Viewing Results
Related Topics
•
Using the Predictive Planning Panel
•
Using Comparison Views
Using the Predictive Planning Panel
When you run a prediction in Predictive Planning, results are displayed in the
Predictive Planning panel. These results are primarily used to compare Predictive
Planning predictions with planners' forecasts. They can also be used to compare other
types of predictions as well as values for various historical time series.
Initially, a chart is displayed. You can also view data or statistics. For all views, the
Member list determines which member is displayed. If you predicted results for more
than one member, look at all results by selecting each member in the list. After you
select a member, you can use the arrow keys to scroll up and down through the
member list.
Note:
Results charts are also called comparison views. For more information about
displaying, editing, and creating them, see Using Comparison Views.
If available, the Pin Panel button,
, detaches the pane from the side panel. You
can move the panel around the screen. Click the Pin Panel button again to attach it
back to the side.
Note:
If the Predictive Planning panel is hidden, select Panel in the Smart View
ribbon to display it again.
You can click the Help button,
, to display online help.
In the Comments panel below the displayed results, you can click the Pivot button,
, to move the Comments panel to the right of the results. Click again to move it
back.
2-1
Chapter 2
Using the Predictive Planning Panel
Chart Tab
Predictive Planning results are displayed graphically on the Chart tab (Figure 1).
Figure 2-1
Predictive Planning Panel, Chart Tab with Summary Area
The default view, Prediction, includes plots of historical and predicted data. The
historical data series is displayed to the left of the vertical separator line. The predicted
data series is bounded by dotted lines that show the upper and lower confidence
intervals (labeled Worst Case and Best Case).
To change the appearance of a chart, double-click it or click the Chart Preferences
button,
(Setting Chart Preferences).
2-2
Chapter 2
Using the Predictive Planning Panel
You can use the Chart Scale button,
, to display a slider control that enables you
to show more or less detail in the chart. You can also display a prediction fit line, a
trend line (best fitting line), a growth rate line, or other scenario data from the
application (Editing the Current View).
You can click the Adjust Series button,
(Adjusting Future Data Series).
, to change values in future data series
Data Tab
The Data tab shows a column for each data series displayed on the chart for the
selected members (Figure 1). In the default display, columns for the Worst Case and
Best Case data series are also included. As in the Chart tab, the Data tab is split into
past and future data sections. The future data section is shown at the bottom of the
data table in bold font.
Note:
Data values in the past section of the Prediction column are plotted as the
prediction fit line when that data series is selected as part of editing a
comparison view (Adding Prediction Data).
2-3
Chapter 2
Using the Predictive Planning Panel
Figure 2-2
Predictive Planning Panel, Data Tab
Statistics Tab
The Statistics tab shows several statistics about historical data used to generate the
prediction: number of values, minimum value, mean value, maximum value, standard
deviation, and the period of seasonality if present (Figure 1).
•
Number of data values—The number of historical data values in the date range
•
Minimum—The smallest value in the date range
•
Mean—The average of a set of values, found by adding the values and dividing
their sum by the number of values
2-4
Chapter 2
Using the Predictive Planning Panel
•
Maximum—The largest value in the data range
•
Standard deviation—The square root of the variance for a distribution, where
variance measures the degree of difference of values from the mean
•
Seasonality—Whether the data has a detectable pattern (cycle) and, if so, the
time period of that cycle
Figure 2-3
Predictive Planning Panel, Statistics Tab
The table also displays the following:
•
An accuracy value
•
The current error measure used to select the best time-series forecasting method
(the default is root mean squared error, RMSE); see Time-series Forecasting Error
Measures for a list.
•
The name of the best time-series forecasting method (Classic Time-series
Forecasting, ARIMA Time-series Forecasting Methods)
2-5
Chapter 2
Using the Predictive Planning Panel
•
The parameters for that method (Classic Nonseasonal Forecasting Method
Parameters, Classic Seasonal Forecasting Method Parameters)
For more information about prediction accuracy, see Summary Area.
Summary Area
By default, the Summary Area is displayed below the results chart or table. It
indicates whether the prediction was successful or whether a warning or error
condition occurred instead. The Summary Area can be used with the Filter Results
feature (Filtering Results) to provide a quick overview of the status of the prediction. If
the prediction succeeded, a quality rating is displayed (see About Prediction Accuracy
later in this topic for details). If results are filtered, messages indicate the filtering
criteria currently in effect.
About Prediction Accuracy
Statistically, the accuracy value is the average percentage error over the entire
prediction period. Accuracy ranges from 0 to 100% and is about 90% in the illustrated
example (Figure 1). Ratings of 95 to 100% are considered Very Good, 90 to 95% are
considered good, 80 to 90% are considered Fair, and 0 to 80% are considered Poor.
Notice that these ratings do not indicate whether the results of the member prediction
are good or not within a planning context, only whether the accuracy of the prediction
is good or not.
Prediction accuracy is a relative measure that considers the magnitude of the
prediction errors in relation to the range of the data. For example, in some cases, the
historical data may appear "noisy" and have apparently large prediction errors, but the
accuracy may still be considered good, because the peaks and valleys of the data and
the size of the prediction errors are small relative to the entire range of the data from
minimum to maximum values.
Setting Chart Preferences
To change the appearance of a chart in the Predictive Planning panel:
1.
Double-click the chart or click the Chart Preferences button,
2.
Select appropriate settings in the Chart Preferences dialog.
3.
Optional: Select Reset to restore default settings.
4.
Select OK when settings are complete.
.
Chart Preferences dialog settings are as follows, when selected:
•
Highlight seasonality—Uses vertical bands to separate periods of cyclical data
(years, months, and so on)
•
Highlight missing values and outliers—Graphically emphasizes filled-in or
adjusted-outlier data if these are present
•
Show separator between past and future data—Displays a vertical line between
historical and predicted data sections
•
Show current view name in chart—Displays the name of the current view in the
results chart
2-6
Chapter 2
Using the Predictive Planning Panel
•
3D chart—Adds a depth perspective to the chart without actually adding a third
measured dimension
•
Transparency—Reduces the intensity of chart colors by the indicated percentage
to better show gridlines or other marks in charted areas
•
Gridlines—Indicates whether lines should be displayed in the chart background,
and, if so, whether they should be vertical, horizontal, or both.
•
Legend—Indicates whether a chart legend should be displayed, and, if so,
whether it should be located to the right of, to the left of, or at the bottom of the
chart, or whether the location should be automatically selected depending on
panel size and orientation
Note:
Changing these settings affects only the appearance of charts on your local
computer and does not affect the charts of other users.
Adjusting Future Data Series
Prediction charts typically show actual data followed by future series such as predicted
values and "worst case/best case " values (Figure 1). You can adjust any future series
by activating a "chart grabber" and manipulating charted data with the mouse or by
using the Adjust Series dialog. When you release the mouse or click OK in the dialog,
changes are immediately pasted to the matching series on the form.
Adjusting Future Series with the Mouse
To adjust future series with the mouse:
1.
Begin by clicking the future data series, either the main prediction line or one of
the bounds, such as Worst Case and Best Case.
This activates the chart grabber (Figure 1). An x is displayed for eachy data point
and a triangle, the chart grabber, appears at the end of the line.
By default, the data points are "unlocked" and can be adjusted evenly.
2-7
Chapter 2
Using the Predictive Planning Panel
Figure 2-4
2.
Clicking the Prediction Line to Activate the Chart Grabber
Perform one of the following actions:
•
Move the chart grabber up or down to increase or decrease all values evenly
with the first period value unlocked (Figure 2).
Figure 2-5
•
Lowering the Chart Grabber Decreases All Values Equally
Click a predicted data point and move it to adjust only that value (Figure 3). A
tooltip indicates which value is adjusted and how it is changing.
2-8
Chapter 2
Using the Predictive Planning Panel
Figure 2-6
•
Moving a Single Data Point
Right-click and select Lock First Period to keep the first predicted value
constant. Move the chart grabber up or down to increase or decrease all
values relative to the first value (Figure 4).
Note:
For more information about locking, see Using the Adjust Series
Dialog.
Figure 2-7
Locked
3.
Moving the Chart Grabber with the First Predicted Value
You can right-click and select Reset at any time to restore the original predicted
values. Otherwise, the adjusted values replace the original values.
See Using the Adjust Series Dialog to perform the same actions using a dialog instead
of manipulating the chart with the mouse. You can right-click and select Adjust Series
to display the dialog from within a chart.
Using the Adjust Series Dialog
To adjust predicted values using a dialog instead of the mouse:
1.
In a Predictive Planning chart, click the Adjust Series button,
.
2-9
Chapter 2
Using Comparison Views
2.
In the Adjust Series dialog, use the Selected series menu to select a predicted
series to adjust.
3.
Select one or more adjustments:
•
Adjust values by — Specify the amount to adjust all values in the selected
series.
•
Round values to — Select No Rounding or a rounding level: Integers, Tens,
Hundreds, Thousands, or Custom.
For Custom, enter a number from -15 to 15 to indicate the rounding level: 0 =
first place to the left of the decimal (units place), 1 = second place to the left of
the decimal (tens), 2 = third place to the left of the decimal (hundreds), 3 =
fourth place to the left of the decimal (thousands), –1 = first place to the right
of the decimal (tenths), –2 = second place to the right of the decimal
(hundredths), –3 = third place to the right of the decimal (thousandths), and so
on. The default level is 0.
•
4.
Restrict values to range — Optionally, enter lower or upper limits for
adjusted values. Defaults are –Infinity to +Infinity.
Optional: Select Lock first period to keep the value of the first predicted value
constant and apply the full set of adjustments to the last predicted value in the
series. Predicted values between those two are scaled accordingly. You can click
to review that definition.
5.
Click OK to perform the adjustment and paste adjusted values into the Planning
form.
6.
Optional: Click Reset to restore the original values for the currently selected
series.
Using Comparison Views
Predictive Planning is shipped with several predefined chart views:
•
Prediction—Includes the historical data series, usually an Actual scenario, and
the predicted future values based on those; the default
•
Scenario 1 vs. Prediction—Compares data for a scenario mapped as Scenario 1
in the Set Up Prediction dialog with the predicted data; does not include the
historical data series
•
Scenario 2 vs. Prediction—Compares data for a second scenario mapped as
Scenario 2 in the Set Up Prediction dialog with the predicted data; does not
include the historical data series
•
Historical Scenario 1 vs. Historical Prediction—Similar to Scenario 1 vs.
Prediction but compares only historical values
•
Historical Scenario 1 vs. Historical Scenario 2—Compares historical values for
two scenarios mapped in the Set Up Prediction dialog
Notice that these predefined views may not be available if the associated scenarios
have not been mapped in the Set Up Prediction dialog.
You can edit predefined or custom views, create new custom views, and manage
views.
2-10
Chapter 2
Using Comparison Views
Note:
Editing or creating views only affects the views on your local computer and
does not affect the views of other users.
Editing the Current View
Note:
You use very similar dialogs to edit the current view and create a new view,
except that you can edit the name of a new view.
To edit the current view:
1.
Select Edit Current View on the Predictive Planning ribbon or right-click the
tabbed portion of the Predictive Planning panel.
Note:
To create a new view, follow the instructions in Creating a New View.
The New View dialog is identical to Edit View.
2.
Select data series to show in the chart and clear the rest.
Each data series in the view can include a Past section, which contains historical
data, and a Future section which contains future predicted values or other
forward-looking values. The point of time that separates the Past and Future
sections is determined when you run a prediction. Prediction items are described
in Adding Prediction Data).
3.
Optional: Use the buttons to add scenarios (Adding a Scenario), prediction data
series (Adding Prediction Data), and trend lines (Adding a Trend Line).
Trend lines can be best fit lines through the historical data or lines based on a
specified percentage of growth.
4.
Optional: Click Remove to delete the selected item from the Data Series list and
the view.
5.
Optional: Use the arrow keys to change the order of selected items in the list, the
lines on the chart, and the columns in the Data tab.
6.
Optional: If you are creating a new view, either accept the automatically
generated name or clear Auto, and then enter a new name in the View Name text
box.
7.
Click OK.
2-11
Chapter 2
Using Comparison Views
Note:
You can use Reset at any time to restore default settings to predefined views
shipped with Predictive Planning.
Adding a Scenario
To add a scenario to a view:
1.
In the Edit View or New View dialog, click Add Scenario.
2.
In the Member Selection dialog, select a member from the Scenario dimension.
3.
Optional: Select a member from the Version dimension, or leave Version
members unselected to use the form's version.
4.
Click OK.
Adding Prediction Data
To add prediction data to a chart view:
1.
In the Edit View or New View dialog, click Add Prediction.
2.
Select from among available prediction data series:
•
Prediction base case—Median prediction values calculated based on past
historical data; median values mean that the actual values in the future are
equally likely to fall above or below the base case values
•
Prediction worst case—A calculated lower confidence interval, by default the
2.5 percentile of the predicted range
•
Prediction best case—A calculated upper confidence interval, by default the
97.5 percentile of the predicted range
•
Prediction fit line—A line of the best fitting time-series forecasting method
through the historical data
If a prediction data series is already in the view, it is checked and not editable. You
can remove the data series by selecting it in the Edit View or New View dialog and
clicking Remove.
3.
Click OK.
Adding a Trend Line
Trend lines on charts can be lines of best fit through historical data or growth rate lines
that increase historical data by a specified percentage.
To add trend lines to a chart:
1.
In the Edit View or New View dialog, click Add Trend Line.
2.
In Add Trend Line, select Linear trend line or Growth rate.
The sample chart shows the effect of your selection.
3.
Optional: If you select Growth rate, specify the rate (2% is the default) and the
time dimension (Year is the default). To compound growth by adding the
2-12
Chapter 2
Using Comparison Views
previously calculated percentage to the current baseline value when calculating
the next value, select Use compounding. By default, this setting is not selected.
4.
Click OK.
Creating a New View
To create a new comparison view:
1.
Select New View on the Predict ribbon.
The New View dialog opens with default settings based on the current view. This
dialog is identical to the Edit View dialog, except that the View Name box is
editable when Auto is cleared and a new view is created when you click OK.
2.
Add or remove data series to create the new view as described in Editing the
Current View.
3.
Because each view must have a unique name, either accept the automatically
generated name or clear Auto and enter a new name.
4.
Click OK to save the new view.
Managing Views
To edit, rename, remove, or reorder any built-in or custom view:
1.
Select Manage Views on the Predictive Planning ribbon.
2.
Select a view on the list and click the appropriate button:
•
Edit opens the Edit View dialog (Editing the Current View).
•
Rename opens the Rename View dialog. Enter a unique name and click OK.
•
Remove deletes the selected view without confirmation.
3.
Optional: Use the arrow buttons to move the selected view to another position in
the list. This changes the order of views in the Comparison Views menu.
4.
Optional: Use the Reset button to restore all predefined views to their default
states.
Warning! Using Reset permanently removes any custom views you created.
5.
Click OK.
2-13
3
Analyzing Results
Related Topics
•
Overview
•
Filtering Results
•
Pasting Results
•
Creating Reports
•
Extracting Data
Overview
You can perform the following tasks to simplify analysis of Predictive Planning results:
•
Filtering Results—Displaying subsets of results
•
Pasting Results—Adding predicted data into Prediction scenarios
•
Creating Reports—Displaying formatted results for selected members
•
Extracting Data—Creating tables of predicted data in Oracle Smart View for Office
Filtering Results
Filtering enables you to display only results that meet certain criteria. For example, you
can set the criteria to show only members that have warning messages. The default is
to show results for all members. When filtering criteria are changed, all open forms are
updated:
•
By default, member rows that do not meet the filtering criteria are collapsed to hide
them. You can change this setting in the General Options dialog (Setting General
Predictive Planning Options).
•
The member list in the Results View is changed to show only members that meet
the filtering criteria and the view is updated.
Note:
Filtering is a global setting. It applies to all forms and persists from one
session to the next. If you save a filtered workbook and reopen it later, you
can display hidden rows by performing a Refresh in Oracle Smart View for
Office.
To filter Predictive Planning results:
1.
In the Predict ribbon, select Filter Results.
3-1
Chapter 3
Pasting Results
2.
In the Filter Results dialog, select a category:
•
Prediction status—The type of icon shown in the comments: Success,
Warning, or Error
•
Prediction accuracy—Determined by a formula based on MAPE (mean
absolute percentage error)
•
Error measure (RMSE, root mean squared error, MAPE, or MAD, mean
absolute deviation)—The error measure to use for selecting the best timeseries forecasting method, specified in the Set Up Prediction dialog.
3.
Select a conditional operator: = (equal to), <> (not equal to), <= (less than or equal
to), >= (greater than or equal to)
4.
Select or enter a value. For Prediction accuracy, values range from 0%-100%;
for Error measure, from 0 to +infinity or 0%-100%, depending on the selected
measure.
5.
Optional: Click Add Row to define another set of selection criteria. Multiple rows
of criteria must all be satisfied to select a member (an AND operation).
6.
Click OK to display members that meet the selected criteria.
Note:
At any time, you can click Reset to remove all selected criteria and
display results without filtering.
Pasting Results
Pasting results enables you to manually copy prediction results into a scenario on the
form, for example a scenario named Forecast.
Tip:
If you want to save prediction data for later comparisons without overwriting
other scenarios, special Prediction scenarios must be added to the form by
an administrator or other user who is able to modify Planning forms before
you use Predictive Planning.
Note:
An administrator or other user who is able to modify Planning forms can map
a scenario to hold base case, best case, or worst case prediction results.
Then, prediction resulots are automatically pasted into that scenario
(Mapping Member Names).
To manually paste prediction results into a scenario on the form:
3-2
Chapter 3
Creating Reports
1.
Determine that a Prediction or other special scenario exists so you will not
overwrite data in other scenarios.
2.
Select one or more members to paste.
Click the bottom half of the Paste Results button and select from the listed
options. (If you click the top half of Paste Results, the Paste Results dialog is
displayed for the current member only; see step 3, following.)
Select from the following:
3.
•
Current Member—Pastes results for only the member that is currently
selected in Results View
•
All Members—Pastes results for all predicted members; if present, filtering is
ignored
•
Filtered Members—When filtering is active, pastes results for the current set
of filtered members
•
Selected Members—Enables you to select members to paste
Select scenarios for pasting in the Paste Results dialog:
•
From—Lists all series in the current view that are available for pasting; select
the one whose data will be copied
•
To—Lists all scenario/version combinations in the form; select the one to
receive the pasted data
•
Prediction range—Select the first setting to use the entire prediction range or
select the second and specify how many periods of data to use
Note:
If the prediction range overlaps the data range on the form, only the
dates shown on the form are pasted.
4.
When settings are complete, click OK.
Creating Reports
Predictive Planning reports can provide several kinds of information about predictions
for selected members, including the run data and time, data attributes, run
preferences, and the prediction results.
To create a Predictive Planning report:
1.
In the Analyze menu or group, select Create Reports.
2.
In the Create Report dialog, select one of the following:
3.
•
All members—Shows report information for all predicted members
•
Filtered members—If available, shows information for all members that are
not excluded by filters
•
Selected members—Displays a dialog for member selection
Optional: Click Preferences to customize the contents of the report (Setting
Report Preferences).
3-3
Chapter 3
Extracting Data
4.
When settings are complete, click OK.
Setting Report Preferences
Creating Reports describes how to generate a basic Predictive Planning report. Report
preferences enable you to customize reports.
To set report preferences:
1.
In the Create Report dialog, click Report Preferences.
2.
On the Report tab of the Report Preferences dialog, in the Report sections list,
select Report Summary to review and, optionally, modify the display selections:
3.
4.
•
Report title—Displays a default report title
•
Run date/time—The date and time the report was created
•
Data attributes—The number of members and other descriptors including the
historical data source
•
Run preferences—The number of periods to predict, whether missing values
are filled in, whether outliers are adjusted, prediction methods used, and the
selected error measure
•
Prediction results—A summary of the predicted values
In the Report sections list, select Members to review and, optionally, modify the
selections:
•
Chart—Includes the results chart at the indicated percent of default size
•
Predicted values—Values for each time period in the prediction range
•
Statistics—Information included in the Statistics tab (Statistics Tab)
•
Methods—The number of time-series forecasting methods reported: all
methods used, the three best methods, the two best methods, or only the best
method, where "best" is defined as the most accurate
On the Options tab of the Report Preferences dialog, review and, optionally,
modify the following settings:
•
Location—Whether to create the report in a new Microsoft Excel workbook or
the current workbook; if you select Current workbook, a new sheet is created
after the current sheet
You can enter a name for the new sheet in the Sheet Name text box.
•
Formatting—Whether to include cell locations (workbook, worksheet, and cell
address) in report headers (selected by default)
•
Chart format—Whether to create a native Predictive Planning chart (Image)
or a Microsoft Excel chart
If you select Image, you can format charts using the Predictive Planning Chart
Preference settings (Setting Chart Preferences).
5.
When all settings are complete, click OK.
Extracting Data
You can extract results and methods from the current Predictive Planning run.
3-4
Chapter 3
Extracting Data
To extract results:
1.
In the Analyze menu or group, select Extract Data.
2.
In the Extract Data dialog, select one of the following:
•
All members—Shows report information for all predicted members
•
Filtered members—If available, shows information for all members that are
not excluded by filters
•
Selected members—Displays the Smart View dialog for member selection
3.
Optional: Click Preferences to select which data to extract (Setting Data
Extraction Preferences).
4.
When settings are complete, click OK.
Setting Data Extraction Preferences
Extracting Data describes how to extract basic Predictive Planning results to a
workbook in tabular form. Data extraction preferences enable you to customize which
results to extract.
To set data extraction preferences:
1.
In the Extract Data dialog, click Preferences.
2.
On the Data tab of the Extract Data Preferences dialog, select the type of data to
extract:
3.
4.
•
Results Table—Extracts past or future values, or both, for the members
selected for data extraction
•
Methods Table—Lists the best time-series forecasting methods plus any of
the following statistical data and information about the forecasting methods
used:
–
Accuracy—An estimate of the quality of predicted results
–
Errors—Error statistics for predicted results (RMSE, MAD, and MAPE)
–
Parameters—Displays calculated parameters for the basic forecasting
methods and transformational lambda and BIC results for ARIMA methods
–
Ranking—Indicates the prediction ranking of displayed methods, where 1
is best
On the Options tab, review and, optionally, modify the following:
•
Location—Indicates whether to write results to a new workbook or the current
workbook, and the sheet names to use for the Results table and Methods
table
•
Formatting—Indicates whether to automatically format results (AutoFormat
selected)
When all settings are complete, click OK.
3-5
4
Setting General Predictive Planning
Options
Setting Up Predictive Planning describes how administrators (and other users whose
security roles enable them to modify Planning forms) can set up Planning and its
Predictive Planning feature for efficient and effective use. This chapter describes how
other users can customize Predictive Planning for individual sessions without
modifying forms.
To change general Predictive Planning option settings:
1.
Select Options in the Predictive Planning menu or ribbon.
2.
Review and, optionally, change General options:
3.
4.
•
Show ribbon only for valid Planning forms—When selected, hides the
Predict ribbon unless a valid form is open; default is selected.
•
Collapse rows and columns on form during filter operations—When
selected, "hides" excluded members by collapsing their rows or columns;
default is selected.
•
Reset Alerts button for "Do not show" checkboxes—When clicked, clears any
checkboxes that were selected to prevent the repetitive display of message
boxes, prompts, and other information where "Do not show" checkboxes are
offered.
Review and, optionally, change Date formatting options:
•
Format—Indicates whether the period or year is displayed first in date labels;
default is Period-Year.
•
Separator— Indicates whether to use -, /, or a blank space to separate the
period and year; default is -.
Optional: Select Enable accessibility options to activate Predictive Planning
features for users with visual impairments, including the use of patterns instead of
colors.
For a description of accessibility features, including keyboard command
equivalents, see the Accessibility Guide for Oracle Planning and Budgeting Cloud
Service.
5.
When settings are complete, click OK.
Note:
You can click Reset at any time to restore default settings.
4-1
A
Setting Up Predictive Planning
Related Topics
•
Before You Begin
•
Setting Form Defaults
Before You Begin
Note:
This appendix is for administrators and other users whose security roles
enable them to modify Planning forms.
Predictive Planning is a Planning feature that works within Oracle Smart View for
Office to predict future results from historical data. It is easy to use but requires some
administrative setup.
This section describes Predictive Planning requirements and explains concepts that
are important when setting Planning form defaults for use with Predictive Planning.
While factory defaults are available, forms should be set up with application defaults at
a minimum; some forms might also require individual defaults.
For most efficient setup, review the topics listed at the beginning of this section first,
and then set application and individual defaults (Setting Form Defaults).
Assigning Security Roles
Predictive Planning users must be assigned roles that enable them to use Planning
and to be an ad hoc user. Roles are assigned using Oracle Identity Management. Only
those with the ability to modify forms can use the Set Up Prediction dialog to define
Predictive Planning defaults.
Hierarchical Data Prediction Issues
Planning data is structured in a hierarchy of levels, from the most general categories to
the most detailed. Knowledge of important concepts in this section will help when
working with the Member Selection dialog box and other setup features.
Comparing Bottom-up, Top-down, and Full Predictions
Full predictions, the default, predict all members on a form without regard to dimension
hierarchies. With this method, Predictive Planning makes no assumption about the
type of aggregation on the form.
A-1
Appendix A
Before You Begin
Bottom-up predictions involve predicting members at the lowest levels of the
dimension hierarchies and optionally rolling up the results to higher level summary
members. This type of prediction requires that historical data be available for the
lowest level members.
Top-down predictions involve predicting members at the summary levels of the
dimension hierarchies and optionally spreading the results down to lower level
members. This type of prediction is useful when historical data is not available for
lower level members, or when top level predictions are being used to drive the results
down to lower members.
Note:
Prediction results between full, bottom-up, and top-down methods should be
close; however, predictions on lower level members are the most accurate
because the individual trends and patterns of the data are preserved in the
prediction process. If you are using top-down or full predictions and want to
preserve the results at the summary-levels, ensure that the Planning
business logic does not aggregate the results from lower level members.
Pasting Results for Predictions
To roll up (or spread down) results, users need to paste the predicted values into the
form, and then submit the form. This recalculates the Planning business logic and
propagates the predicted results accordingly. To simplify the pasting of predicted
values by users, you can set up automatic pasting for the form (Mapping Member
Names).
Caution:
If users will be pasting results, either manually or automatically, a scenario
must be added to the form to hold the pasted results. For example, a
Prediction scenario could be added. Otherwise, the pasted results could
overwrite other scenarios. For more information, see Creating a New
Scenario for Prediction Results.
Aggregating Best and Worst Case Predictions
The best and worst case predictions (by default, the 2.5% and 97.5% percentiles of
the predicted values) are automatically generated. These values can be saved in
Planning, but are not easy to roll up or spread down because of the complexity of their
aggregation. Rolling them up or spreading them down requires custom formulas to be
added to the Planning business logic. While closed-form formulas are available for
addition and subtraction, they do not exist for some cases of aggregation (for example,
division).
A-2
Appendix A
Before You Begin
Historical Data and Prediction Accuracy
The amount of historical data available determines the accuracy of the predictions; the
more data the better. There should be at least twice the amount of historical data as
the number of prediction periods. If not enough historical data is available at the time
of prediction, a warning or error is displayed. Predictive Planning can detect seasonal
patterns in the data and project them into the future (for example, spikes in sales
numbers during holiday seasons). At least two complete cycles of data must be
available to detect seasonality.
In addition, Predictive Planning detects missing values in the historical data, filling
them in with interpolated values, and scans for outlier values, normalizing them to an
acceptable range. If there are too many missing values or outliers in the data to
perform reliable predictions, a warning or error message is displayed.
Form Creation and Modification Issues
Certain aspects of form structure affect the performance of Predictive Planning, as
described in the listed topics.
Using Valid Forms
Before you can predict, make sure you have a valid form. In general, a valid form must
have the following:
•
A series axis, containing one or more non-time dimensions, such as Account or
Entity. The series axis may not contain Year or Period dimensions.
•
A time axis, containing the Year or Period dimensions, or both. The time axis may
contain Scenario and Version dimensions. The time axis may not contain other
non-time dimensions.
•
Scenario and Version dimensions are permitted on the series or time axes, or
both.
•
The form must not be empty.
Determining the Time Granularity of Predictions
The lowest Period dimension member level on a form determines the time granularity
of the prediction. That is, if the lowest member level is Quarters (Qtr1, Qtr2, on so on),
then historical data is retrieved at the Quarters level and the prediction will also take
place at the Quarters level. For this reason, it is important to include on the form the
lowest level of Period members possible so that the greatest amount of historical data
can be used.
In Figure 1, Quarters are the lowest level members of the Period dimension that
appear on the form. You can tell this by the fact that the "Q1" name does not have a
"+" symbol by it. If it did, this would mean that lower level members (such as months)
exist on the form but are hidden from view by collapsing the columns. If the form
included the Months levels (even if hidden), then Predictive Planning would predict at
the Months level. For purposes of determining time granularity, it does not matter if the
members are hidden or visible on the form.
A-3
Appendix A
Before You Begin
Figure A-1
Time-granularity Example
Determining the Prediction Range
The prediction range starts one period after the end of historical data for all members
on the form, regardless of the starting date of the form. If the members do not all have
the same amounts of historical data, the end of historical data (and thus the start of the
prediction range) will be determined by those members that have the greatest amount
of similar historical data. These dates can be overridden by the user at the start of a
prediction. By default, the end date on the form determines the end date of the
prediction. This can also be overridden by users at the start of a prediction.
Note:
The prediction range end date is also limited to the members defined for
Year and Period. That is, if the last Year-Period defined is 2015-Dec, then it
is not possible to predict past this date. This limit is independent of the end
date on the form itself. If users are having trouble predicting too far into the
future and are receiving error messages, more time periods must be defined
in the Planning application.
Creating a New Scenario for Prediction Results
After a prediction runs, users can paste the results to a form and save them. Typically,
users may want to save prediction results to a Forecast or Plan scenario. However, if
A-4
Appendix A
Setting Form Defaults
users want to keep the prediction results separate from these types of scenarios, you
will need to add to add a special scenario to Planning (for example, “Prediction”) to
hold those results without overwriting other scenarios. You can also create additional
scenarios to store the best and worse case prediction results as well. These scenarios
should then be mapped appropriately in the Set Up Prediction dialog (Mapping
Member Names). For additional discussion, see Pasting Results for Predictions and
Aggregating Best and Worst Case Predictions.
Note:
Members that are read-only on the form can still be predicted, but the results
cannot be pasted back into the member rows or columns.
Setting Form Defaults
Setting up a form for use with Predictive Planning defines application or individual
defaults for that form. Some of the settings require Planning knowledge, while others
require a basic knowledge of classic and ARIMA time-series forecasting. Once a form
has been set up, users should be able to open the form in Oracle Smart View for
Office, start Predictive Planning, and immediately run a prediction using the defaults.
Tip:
If other defaults are not available, factory defaults are applied to all forms
used with Predictive Planning. If customized defaults are required,
application defaults can automate that process at an application level, while
individual defaults override other defaults on a particular form. For best
results, read this entire section, particularly Application and Individual Form
Defaults, before setting any Predictive Planning defaults.
Note:
You must have a security role that enables you to modify Planning forms to
define defaults.
Application and Individual Form Defaults
When a form is first opened in Predictive Planning, it receives factory defaults for all
Predictive Planning settings (that is, all of the settings that appear in the Set Up
Prediction dialog). You will probably want to override some of these settings and
create an application-level default for all forms, or individually customize the default
settings for selected forms. The application default settings are stored in the Planning
application and are applied to all forms when they are opened. Individual defaults are
stored with the form to which they are applied.
A-5
Appendix A
Setting Form Defaults
Tip:
Set the application-level default for all forms first, and then customize the
default for individual forms as needed.
To set application-level defaults:
1.
Open any form.
2.
Customize the settings in the Set Up Prediction dialog.
3.
Click Set Default.
All settings on all tabs of the Set Up Prediction dialog are immediately saved as
application defaults for all forms.
4.
Press Cancel to avoid setting an individual-level default for the current form.
To set individual-level defaults:
1.
Open a form and customize the settings in the Set Up Prediction dialog.
2.
Click OK to save all settings on all tabs as individual defaults.
Whenever that form is opened, all the settings are applied and override any
application-level defaults.
When forms are opened by users, the form first receives any individual-level default
settings, if an individual default was created, and then receives application-level
defaults.
Using the Set Up Prediction Dialog
The Set Up Prediction dialog is used to do the following:
•
Select the source of historical data on which to base predictions (Specifying a
Historical Data Source)
•
Map Predictive Planning names to members (Mapping Member Names)
•
Specify which members on a form to predict (Selecting Members)
•
Select and override various prediction option settings (Setting Prediction Options)
To open the Set Up Prediction dialog, select Set Up Prediction,
Predictive Planning ribbon.
, in the
Specifying a Historical Data Source
When you specify a historical data source, you select where the historical data will be
coming from and indicate whether to use all historical data or only data from a
specified date range.
A-6
Appendix A
Setting Form Defaults
Note:
Administrators and other users with appropriate security roles can define and
use alternate data sources instead of or in addition to the default data source
for the current Planning application (Using Alternate Historical Data
Sources).
To specify a source for historical data:
1.
Open the Set Up Prediction dialog.
2.
On the Data Source page, select a Plan Type:
3.
•
PlanName (Default Plan) is the Plan Type associated with the current form.
Select this plan type to use any historical data contained within this application
(the default).
•
OtherPlanNames, if available, are alternate plan types provided by the data
administrator as sources of historical data. These are typically Aggregate
Storage Option (ASO) applications.
Indicate whether to Use all historical data or a Selected date range.
Note:
When they run predictions, users will be able to temporarily override the
selected date range using the Change Date buttons on the Run
Confirmation dialog.
4.
Optional: If you selected Selected date range, specify a start and end year and
time period.
Note:
For a discussion of the date range, see Determining the Prediction
Range.
5.
Optional: Set or reset defaults using one of the following selections:
•
Click Set Default to store settings on all tabs as application defaults.
•
Click OK to store settings on all tabs as individual defaults for only this form.
•
Click Reset at any time to restore the predefined defaults shipped with
Predictive Planning or application defaults set with Set Default. This resets all
tabs of the dialog.
A-7
Appendix A
Setting Form Defaults
Note:
For more information about defaults, see Application and Individual
Form Defaults.
6.
Optional: To leave the dialog without changing defaults, click Cancel.
Mapping Member Names
Use Map Names to identify key scenarios in the application and link them to Predictive
Planning data series. Predictive Planning uses the historical data series to generate
predictions for each member on the form. Comparison data series can be set up to
compare predicted results to forecast scenarios, budget scenarios, and so on.
Prediction data series can be set up to hold prediction results in a separate area in the
application. For details, see About Name Defaults.
To map member names to specific Predictive Planning data series:
1.
Open the Set Up Prediction dialog.
2.
On Map Names, select the following:
•
Historical data series group, Scenario—The dimension member name to
use as the historical data series to generate the prediction; a required
selection
•
Comparison data series group, Scenario 1 and Scenario 2—Additional
dimension member names to compare with the historical data series in
comparison charts; selecting one or both scenarios in this group is optional
•
Prediction data series group, Base case scenario, Worst case scenario,
and Best case scenario—Optional scenarios that must be created in
Planning forms by administrators or other users whose security roles enable
them to modify Planning forms; used to hold predicted values when pasted
into the form
To select a member, click the ... button, and then select members from the
Scenario and Version dimensions. If you do not select a Version member, the
current Version member on the form is used. If there are more than one Version
members on the form, the first Version member is used.
3.
Optional: When a Comparison data series or Prediction data series member
is selected, an X button is displayed next to it. You can use this button to clear the
selection and restore the list to its default, <None>.
Because the Historical data series member is required, you can not clear it and
can only select another member.
4.
Optional: Set or reset defaults using one of the following selections:
•
Click Set Default to store settings on all tabs as application defaults.
•
Click OK to store settings on all tabs as individual defaults for only this form.
•
Click Reset at any time to restore the predefined defaults shipped with
Predictive Planning or application defaults set with Set Default. This resets all
tabs of the dialog.
A-8
Appendix A
Setting Form Defaults
Note:
For more information about defaults, see Application and Individual
Form Defaults.
5.
Optional: To leave the dialog without changing defaults, click Cancel.
About Name Defaults
The Map Names panel on the Set Up Prediction dialog is used to identify Predictive
Planning key scenarios on the form. The only required mapping identifies which
scenario holds the historical data series; the default is “Actual ([current])”. You will
need to change this default if the historical data scenario is something other than
“Actual”, or if the version for this scenario is different from the form’s version. To make
it easier for users to compare predicted results to other scenarios like Forecast or
Plan, you can map these scenarios in the Comparison data series section.
When users open the form, several additional views automatically appear in the
Comparison Views menu, and users can select from among these comparisons. If you
do not map the comparison data series, users can always create custom comparison
views manually using the Edit Current View and New View commands. Manually
created views are stored only on the user’s computer. If you add special scenarios to
Planning to hold prediction results, you should map these scenarios in the Prediction
data series section. For instructions, see Mapping Member Names.
Selecting Members
Use Member Selection to determine which form members to select for prediction. Full
predictions, the default, choose all members on the form. "Bottom-up" predictions
choose members at the lowest level of the hierarchy for forms built to aggregate
results up to higher level members. "Top-down" predictions choose members at the
highest level of the hierarchy for forms built to push results down to lower level
members. Optionally, you can skip any read-only members.
Note:
When running predictions, users can override these settings using the
Change Member Selection button on the Run Confirmation dialog. Its
settings are similar to the following but they apply only temporarily to the
current Predictive Planning session.
To indicate which members on a form to include in a prediction:
1.
Open the Set Up Prediction dialog.
2.
On Member Selection, select a prediction type:
•
Bottom-up (lowest level members only)—Includes only the lowest level
members in the hierarchy included on the form, the lowest level for each
dimension if multiple dimensions are included
A-9
Appendix A
Setting Form Defaults
•
Top-down (highest level members only)—Includes only the highest level
members in the hierarchy included on the form, the highest level for each
dimension if multiple dimensions are included
•
Full (all members)—Predicts all members regardless of their hierarchy level;
the default
3.
Optional: Select Skip 'read only' members, which includes only members with
writable (editable) cells in the prediction. Members with read-only cells typically
include calculated summary data that is stored in the dimension hierarchy.
4.
Optional: Set or reset defaults using one of the following selections:
•
Click Set Default to store settings on all tabs as application defaults.
•
Click OK to store settings on all tabs as individual defaults for only this form.
•
Click Reset at any time to restore the predefined defaults shipped with
Predictive Planning or application defaults set with Set Default. This resets all
tabs of the dialog.
Note:
For more information about defaults, see Application and Individual
Form Defaults.
5.
Optional: To leave the dialog without changing defaults, click Cancel.
Setting Prediction Options
The prediction options specify data attributes, prediction methods, and other aspects
of time-series analysis performed by Predictive Planning. The defaults are suitable for
most predictions and should only be changed by those with some knowledge of timeseries analysis.
To set prediction options:
1.
Open the Set Up Prediction dialog.
2.
On Options, review and select from the following:
•
Data attributes group:
–
Select whether to detect seasonality (regular cycles of data) automatically
(Automatic, the default) or manually (Manual). If you select Manual,
specify the number of time periods per cycle For example if time periods
are quarters with a yearly cycle, there would be 4 periods per cycle.
–
Select whether to Fill in missing values and Adjust outliers. These
settings estimate missing data based on adjacent data and help to
normalize unusual data.
A-10
Appendix A
Setting Form Defaults
Note:
Fill-in Missing Values uses interpolation to fill in gaps in the
historical data. Clearing this option skips prediction calculation
for members with gaps in their data.
Adjust Outliers uses a special fitting algorithm to determine
whether data points fall within a reasonable range compared to
all the other data points for a member. Clearing this option still
allows the prediction to proceed, although the prediction
algorithm may be thrown off by the outlier data points.
•
Prediction methods group:
–
Select which time-series prediction methods to use: Nonseasonal (does
not fit to cyclical data), Seasonal (fits to cyclical data), or ARIMA (both
nonseasonal and seasonal using predefined statistical models). See
Classic Time-series Forecasting and ARIMA Time-series Forecasting
Methods for lists and details.
Select all three, the default, unless you have a good reason to do
otherwise.
–
Select an error measure to use in selecting the best method: RMSE,
MAD, or MAPE (Time-series Forecasting Error Measures).
Again, use the default, RMSE, unless you have a good reason to use
another.
•
Prediction periods group:
–
Select whether to detect periods automatically, Select periods based on
form, or manually, Manual. If you select Manual, specify the number of
periods to predict. Generally, the number of prediction periods should be
less than half the amount of actual data.
–
Select a Prediction interval, which defines a range around the base
predicted value where the value has some probability of occurring; for
example, the default (2.5% and 97.5%) means that there is a 95%
probability that the predicted value will fall between the 2.5 percentile and
the 97.5 percentile.
A-11
Appendix A
Setting Form Defaults
Note:
Prediction Interval determines the percentile range around the
base case prediction that is used to represent the best and worst
case predictions. For example, a 2.5% - 97.5% prediction
interval estimates that 95% of the time the predicted value will
actually occur between the lower and upper bounds; 5% of the
time the value will lie outside of these bounds.
These lower and upper percentile values are also used to
indicate the worst and best case predicted values. For a
Revenue-type account member, the worst and best cases are
assigned to the lower and upper percentile values, respectively.
For an Expense-type account member, the cases are reversed;
the best case is associated with the lower bound (e.g. 2.5%) and
the worst case is associated with the upper bound (e.g. 97%).
3.
Optional: Set or reset defaults using one of the following selections:
•
Click Set Default to store settings on all tabs as application defaults.
•
Click OK to store settings on all tabs as individual defaults for only this form.
•
Click Reset at any time to restore the predefined defaults shipped with
Predictive Planning or application defaults set with Set Default. This resets all
tabs of the dialog.
Note:
For more information about defaults, see Application and Individual
Form Defaults.
4.
Optional: To leave the dialog without changing defaults, click Cancel.
Using Alternate Historical Data Sources
Specifying a Historical Data Source describes how to specify a source for the historical
data used to predict future results. You select the source in the Plan Type box.
The default plan type is the plan associated with the current form, but administrators
and others with appropriate security roles can define and use alternate plan types as
historical data sources. For example, an administrator can create an ASO Plan Type
for historical data, since this type supports efficient storage and access to large
amounts of data (Alternate Plan Types and Dates).
Note:
Alternate plan types can contain data for dates that are earlier than those
included in the default plan type (Alternate Plan Types and Dates).
A-12
Appendix A
Setting Form Defaults
If alternate plans types are available, you can select them for use in the Data Source
panel. If you select an alternate plan type, the upper part of the Data Source panel
includes additional controls:
•
Configure POV button—Opens the Member Selection dialog, where you can add
members that are unmatched in the alternate plan type point of view (POV). See
Alternate Plan Types and POV Configuration.
•
Warning icon—Clicking this icon,
, or pressing the spacebar while it is
selected displays a detailed message about POV issues to help identify
unmatched members for configuration.
•
Consolidate with default plan type checkbox—When selected, this setting
indicates that historical data is taken from the alternate plan type first and then
from the default plan type.
With consolidation, data overlaps or gaps are evaluated for each data series. In
case of overlap, data from the two data sources are merged. Data from the
alternate plan type overrides any data from the default plan type for the same date
location. If there is a gap between the data sets, missing values are estimated and
filled in when a prediction is run.
When Consolidate with default plan type is not selected, historical data is read
only from the alternate plan type.
Alternate Plan Types and POV Configuration
If the point of view for the current form can not be matched to the alternate plan type,
an error message and a warning icon are displayed. You can click the icon to learn
more about the detected mismatch. For example, a member in the POV may not be
present in the alternate plan type and must be configured.
To configure the POV:
1.
Click Configure POV.
2.
In the Member Selection dialog, locate the unmatched member in the first panel
from the left.
3.
Select the value to add and then click the right-arrow in the center of the screen to
move it into the second panel.
4.
When all unmatched members have values, click OK.
Alternate Plan Types and Dates
On reason for defining and using alternate plan types is to enable the use of historical
date ranges that are earlier than those in the default plan type.
The historical data source, whether default or alternate, must contain all the
dimensions on either the Series or Time axis of the current Planning form. One
exception is that an alternate year dimension can be specified for the Year dimension.
This is useful when an alternate plan type contains earlier dates than the default.
About Alternate Year Dimensions
An alternate year dimension may be used for a historical plan type which contains
years prior to the start of the current Year dimension. This approach enables the
A-13
Appendix A
Setting Form Defaults
addition of past historical years if the current Planning application's Year dimension
does not include enough past years to meet prediction requirements. For example, if
the current Year dimension covers FY08 to FY14, it may be necessary to add
historical data from FY03 to FY07 for predictions. In this case, a historical plan type
may be used with an alternate year dimension that contains members FY03 to FY07.
The dimension name may be any valid custom dimension name, such as AltYear. For
dimension requirements, see Alternate Year Dimension Requirements.
Alternate Year Dimension Requirements
Alternate year dimensions must meet the following requirements:
•
The alternate year dimension is a custom Planning dimension with year members
that follow the same naming pattern as the current Year dimension. For example,
if the Year dimension contains FY08 to FY14, then the alternate year dimension
should use FYxx as the naming pattern, such as FY03 to FY07.
•
The application's Year dimension cannot be included in this alternate historical
plan type.
•
When an alternate plan type is selected as a data source and an alternate year
dimension is present, the alternate year dimension will be detected automatically.
A dialog is displayed that asks users if they want to use the alternate year
dimension. If they respond OK, the alternate year dimension is used.
For additional information about creating alternate plan types, see About Creating
Alternate Plan Types
About Creating Alternate Plan Types
Alternate plan types containing alternate year dimensions typically are created after
the initial creation of a Planning application. They usually use the ASO storage type as
this type is more efficient for large amounts of data. All plan types created during the
initial Planning application creation typically inherit the Year dimension. However, ASO
plan types created after the application enable administrators and others with
appropriate security roles to add dimensions selectively so that a custom year
dimension can be included without the default Year dimension.
A-14
B
Predictive Planning Forecasting and
Statistical Descriptions
The topics in this section are for those who want to know more about the forecasting
methods and error measures used in Predictive Planning.
Classic Time-series Forecasting
Two primary techniques of classic time-series forecasting are used in Predictive
Planning:
•
Classic Nonseasonal Forecasting Methods — Estimate a trend by removing
extreme data and reducing data randomness
•
Classic Seasonal Forecasting Methods — Combine forecasting data with an
adjustment for seasonal behavior
For information about autoregressive integrated moving average (ARIMA) time-series
forecasting, see ARIMA Time-series Forecasting Methods.
Classic Nonseasonal Forecasting Methods
Nonseasonal methods attempt to forecast by removing extreme changes in past data
where repeating cycles of data values are not present.
Single Moving Average (SMA)
Smooths historical data by averaging the last several periods and projecting the last
average value forward.
This method is best for volatile data with no trend or seasonality. It results in a straight,
flat-line forecast.
Figure B-1
Typical Single Moving Average Data, Fit, and Forecast Line
B-1
Appendix B
Classic Time-series Forecasting
Double Moving Average (DMA)
Applies the moving average technique twice, once to the original data and then to the
resulting single moving average data. This method then uses both sets of smoothed
data to project forward.
This method is best for historical data with a trend but no seasonality. It results in a
straight, sloped-line forecast.
Figure B-2
Typical Double Moving Average Data, Fit, and Forecast Line
Single Exponential Smoothing (SES)
Weights all of the past data with exponentially decreasing weights going into the past.
In other words, usually the more recent data has greater weight. Weighting in this way
largely overcomes the limitations of moving averages or percentage change methods.
This method, which results in a straight, flat-line forecast is best for volatile data with
no trend or seasonality.
Figure B-3
Typical Single Exponential Smoothing Data, Fit, and Forecast Line
B-2
Appendix B
Classic Time-series Forecasting
Double Exponential Smoothing (DES)
Applies SES twice, once to the original data and then to the resulting SES data.
Predictive Planning uses Holt’s method for double exponential smoothing, which can
use a different parameter for the second application of the SES equation.
This method is best for data with a trend but no seasonality. It results in a straight,
sloped-line forecast.
Figure B-4
Typical Double Exponential Smoothing Data, Fit, and Forecast Line
Damped Trend Smoothing (DTS) Nonseasonal Method
Applies exponential smoothing twice, similar to double exponential smoothing.
However, the trend component curve is damped (flattens over time) instead of being
linear. This method is best for data with a trend but no seasonality.
Figure B-5
Typical Damped Trend Smoothing Data, Fit, and Forecast Line
Classic Nonseasonal Forecasting Method Parameters
The classic nonseasonal methods use several forecasting parameters. For the moving
average methods, the formulas use one parameter, period. When performing a moving
average, Predictive Planning averages over a number of periods. For single moving
B-3
Appendix B
Classic Time-series Forecasting
average, the number of periods can be any whole number between 1 and half the
number of data points. For double moving average, the number of periods can be any
whole number between 2 and one-third the number of data points.
Single exponential smoothing has one parameter: alpha. Alpha (a) is the smoothing
constant. The value of alpha can be any number between 0 and 1, not inclusive.
Double exponential smoothing has two parameters: alpha and beta. Alpha is the same
smoothing constant as described above for single exponential smoothing. Beta (b) is
also a smoothing constant exactly like alpha except that it is used during second
smoothing. The value of beta can be any number between 0 and 1, not inclusive.
Damped trend smoothing has three parameters: alpha, beta, and phi (all between 0
and 1, not inclusive).
Classic Seasonal Forecasting Methods
Seasonal forecasting methods extend the nonseasonal forecasting methods by adding
an additional component to capture the seasonal behavior of the data.
Seasonal Additive
Calculates a seasonal index for historical data that does not have a trend. The method
produces exponentially smoothed values for the level of the forecast and the seasonal
adjustment to the forecast. The seasonal adjustment is added to the forecasted level,
producing the seasonal additive forecast.
This method is best for data without trend but with seasonality that does not increase
over time. It results in a curved forecast that reproduces the seasonal changes in the
data.
Figure B-6
Trend
Typical Seasonal Additive Data, Fit, and Forecast Curve without
Seasonal Multiplicative
Calculates a seasonal index for historical data that does not have a trend. The method
produces exponentially smoothed values for the level of the forecast and the seasonal
adjustment to the forecast. The seasonal adjustment is multiplied by the forecasted
level, producing the seasonal multiplicative forecast.
B-4
Appendix B
Classic Time-series Forecasting
This method is best for data without trend but with seasonality that increases or
decreases over time. It results in a curved forecast that reproduces the seasonal
changes in the data.
Figure B-7 Typical Seasonal Multiplicative Data, Fit, and Forecast Curve
without Trend
Holt-Winters’ Additive
Is an extension of Holt's exponential smoothing that captures seasonality. This method
produces exponentially smoothed values for the level of the forecast, the trend of the
forecast, and the seasonal adjustment to the forecast. This seasonal additive method
adds the seasonality factor to the trended forecast, producing the Holt-Winters’
additive forecast.
This method is best for data with trend and seasonality that does not increase over
time. It results in a curved forecast that shows the seasonal changes in the data.
Figure B-8
Typical Holt-Winters’ Additive Data, Fit, and Forecast Curve
Holt-Winters’ Multiplicative
Is similar to the Holt-Winters’ additive method. Holt-Winters’ Multiplicative method also
calculates exponentially smoothed values for level, trend, and seasonal adjustment to
B-5
Appendix B
Classic Time-series Forecasting
the forecast. This seasonal multiplicative method multiplies the trended forecast by the
seasonality, producing the Holt-Winters’ multiplicative forecast.
This method is best for data with trend and with seasonality that increases over time. It
results in a curved forecast that reproduces the seasonal changes in the data.
Figure B-9
Typical Holt-Winters’ Multiplicative Data, Fit, and Forecast Curve
Damped Trend Additive Seasonal Method
Separates a data series into seasonality, damped trend, and level; projects each
forward; and reassembles them into a forecast in an additive manner.
This method is best for data with a trend and with seasonality. It results in a curved
forecast that flattens over time and reproduces the seasonal cycles.
Figure B-10
Typical Damped Trend Additive Data, Fit, and Forecast Curve
Damped Trend Multiplicative Seasonal Method
Separates a data series into seasonality, damped trend, and level; projects each
forward; and reassembles them into a forecast in a multiplicative manner.
This method is best for data with a trend and with seasonality. It results in a curved
forecast that flattens over time and reproduces the seasonal cycles.
B-6
Appendix B
ARIMA Time-series Forecasting Methods
Figure B-11
Typical Damped Trend Multiplicative Data, Fit, and Forecast Curve
Classic Seasonal Forecasting Method Parameters
The seasonal forecast methods use the following parameters:
•
alpha (α) — Smoothing parameter for the level component of the forecast. The
value of alpha can be any number between 0 and 1, not inclusive.
•
beta (β) — Smoothing parameter for the trend component of the forecast. The
value of beta can be any number between 0 and 1, not inclusive.
•
gamma (γ) — Smoothing parameter for the seasonality component of the forecast.
The value of gamma can be any number between 0 and 1, not inclusive.
•
phi (Φ) — Damping parameter; any number between 0 and 1, not inclusive.
Each seasonal forecasting method uses some or all of these parameters, depending
on the forecasting method. For example, the seasonal additive forecasting method
does not account for trend, so it does not use the beta parameter.
The damped trend methods use phi in addition to the other three.
ARIMA Time-series Forecasting Methods
Autoregressive integrated moving average (ARIMA) forecasting methods were
popularized by G. E. P. Box and G. M. Jenkins in the 1970s. These techniques, often
called the Box-Jenkins forecasting methodology, have the following steps:
1.
Model identification and selection
2.
Estimation of autoregressive (AR), integration or differencing (I), and moving
average (MA) parameters
3.
Model checking
ARIMA is a univariate process. Current values of a data series are correlated with past
values in the same series to produce the AR component, also known as p. Current
values of a random error term are correlated with past values to produce the MA
component, q. Mean and variance values of current and past data are assumed to be
stationary, unchanged over time. If necessary, an I component (symbolized by d) is
added to correct for a lack of stationarity through differencing.
B-7
Appendix B
Time-series Forecasting Error Measures
In a nonseasonal ARIMA(p,d,q) model, p indicates the number or order of AR terms, d
indicates the number or order of differences, and q indicates the number or order of
MA terms. The p, d, and q parameters are integers equal to or greater than 0.
Cyclical or seasonal data values are indicated by a seasonal ARIMA model of the
format:
SARIMA(p,d,q)(P,D,Q)(t)
The second group of parameters in parentheses are the seasonal values. Seasonal
ARIMA models consider the number of time periods in a cycle. For a year, the number
of time periods (t) is 12.
Note:
In Predictive Planning charts, tables, and reports, seasonal ARIMA models
do not include the (t) component, although it is still used in calculations.
Predictive Planning ARIMA models do not fit to constant datasets or datasets
that can be transformed to constant datasets by nonseasonal or seasonal
differencing. Because of that feature, all constant series, or series with
absolute regularity such as data representing a straight line or a saw-tooth
plot, do not return an ARIMA model fit.
Time-series Forecasting Error Measures
One component of every time-series forecast is the data’s random error that is not
explained by the forecast formula or by the trend and seasonal patterns. The error is
measured by fitting points for the time periods with historical data and then comparing
the fitted points to the historical data.
RMSE
RMSE (root mean squared error) is an absolute error measure that squares the
deviations to keep the positive and negative deviations from cancelling out one
another. This measure also tends to exaggerate large errors, which can help eliminate
methods with large errors.
Forecasting Method Selection
All of the nonseasonal forecasting methods and the ARIMA method are run against
the data.
If the data is detected as being seasonal, the seasonal forecasting methods are run
against the data.
The forecasting method with the lowest error measure (for example, RMSE) is used to
forecast the data.
B-8
Download PDF
Similar pages