LWBDID3 001 SolutionsToPractices

Essential Functions of SAS Intelligent Decisioning: Solutions to Practices
Contents
Lesson 1 Solutions ......................................................................................................................................... 2
Solution: Working in SAS Intelligent Decisioning ...................................................................................... 2
Solution: Loading tables into CAS ............................................................................................................. 3
Lesson 2 Solutions ......................................................................................................................................... 3
Solution: Managing and Mapping Variables in a Decision........................................................................ 3
Solution: Creating a Rule Set..................................................................................................................... 5
Solution: Adding a Rule Set to a Decision ................................................................................................. 8
Solution: Creating a Lookup Table .......................................................................................................... 10
Solution: Using Lookup Table Functions in a Rule Set ............................................................................ 10
Solution: Creating and Using a Global Variable ...................................................................................... 12
Solution: Configuring a Branch Node ...................................................................................................... 13
Solution: Configuring a Cross-branch Link .............................................................................................. 14
Solution: Configuring a Record Contacts Node....................................................................................... 15
Solution: Comparing Decision Versions .................................................................................................. 15
Lesson 3 Solutions ....................................................................................................................................... 16
Solution: Creating a Data Query for Use in a Decision ........................................................................... 16
Solution: Adding a Data Query to a Decision .......................................................................................... 18
Solution: Defining DS2 Code for Use in a Decision ................................................................................. 19
Solution: Adding a Model to a Decision .................................................................................................. 20
Solution: Configuring a Data Query to Return a Data Grid ..................................................................... 22
Solution: Using Data Grid Functions ....................................................................................................... 23
Solution: Scoring Rows in a Data Grid..................................................................................................... 26
Solution: Creating and Locking a Treatment........................................................................................... 29
Solution: Creating a Treatment Eligibility Rule ....................................................................................... 31
Solution: Creating a Treatment Group ................................................................................................... 32
Solution: Adding a Treatment Group to a Decision ................................................................................ 33
Solution: Using an Assignment Rule Set for Offer Arbitration................................................................ 35
Solution: Basic Testing of a Decision....................................................................................................... 36
Solution: Scenario Testing....................................................................................................................... 39
Solution: Publishing and Validating a Published Decision ...................................................................... 41
Copyright © 2021, SAS Institute Inc., Cary, North Carolina, USA. ALL RIGHTS RESERVED. 1
Lesson 1 Solutions
Solution: Working in SAS Intelligent Decisioning

In this practice, you log on to SAS Intelligent Decisioning and explore the user interface.
1. Open a browser (such as Google Chrome) in the virtual lab. From the home page, log on to SAS
using the user ID lynn and password Student1.
a. Launch a browser (such as Google Chrome).
b. For User ID, enter lynn.
c. For Password, enter Student1. The password is case-sensitive.
d. Click Sign In.
2. Navigate to SAS Intelligent Decisioning and then to rule sets. Search for rule sets with the text
example in the name.
a. At the top left of the interface, click Show list of applications  Build Decisions.
b. Click Rule sets.
c. In the Search name field at the top left of the list of decisions, enter example and click
Start search.
d. Notice that a rule set named ExampleRuleSet is displayed.
3. Duplicate the rule set named ExampleRuleSet. Name the duplicate DuplicateRuleSet.
a. Enable the check box next to ExampleRuleSet.
b. At the top right of the interface, click Actions  Duplicate.
c. In the Name field, enter DuplicateRuleSet.
d. Click Duplicate.
4. Clear the search text. Sort the rule sets by descending date modified and confirm that
DuplicateRuleSet appears at the top of the list.
a. Click Clear search text.
b. Click the Date Modified column heading in the list of rule sets to sort in ascending
order.
c. Click the Date Modified column heading a second time to sort in descending order.
d. Confirm that DuplicateRuleSet appears at the top of the list.
5. Access the Help Center.

a. Click the L in the circle at the top right of the interface.
b. Click Help Center.
c. The SAS Intelligent Decisioning User's Guide opens in a new browser tab.
6. Sign out of SAS Intelligent Decisioning.

a. Click the SAS Intelligent Decisioning browser tab to return to the application.
b. Click the L in the circle at the top right of the interface.
c. Click Sign out.
Solution: Loading Tables into CAS

In this practice, you load the tables needed for testing in Intelligent Decisioning into
CAS.
Reminder: Sign in to SAS using the user ID Lynn and password Student1.
1. Load the table hmeqapps.sashdat from cas-shared-default  Public into CAS. Hint: At the top
left of the interface, click Show list of applications  Manage data. Then click the Data Sources
tab.
a. At the top left of the interface, click Show list of applications  Manage Data.
b. Click Data Sources.
c. Navigate to cas-shared-default  Public.
d. Click hmeqapps.sashdat.
e. At the top right of the interface, click Load into memory.
Note: Alternatively, you can right-click the SASHDAT file and select Load.
f. Notice that HMEQAPPS appears as a loaded table in the list on the left.
2. Repeat the previous steps to load the tables Investments.sashdat, TSAClaims.sashdat,

SalesReps.sashdat and VendorList.sashdat into CAS.
3. On the Available tab, confirm that all five tables are listed.
a. Click the Available tab.
b. Confirm that the following tables are listed:
 HMEQAPPS
 INVESTMENTS
 SALESREPS
 TSACLAIMS
 VENDORLIST
Lesson 2 Solutions
Solution: Managing and Mapping Variables in a

Decision
In this practice, you create, manage, and map variables in a decision.
Reminder: Sign in to SAS using the user ID Lynn and password Student1. To navigate
to SAS Intelligent Decisioning, click Show list of applications  Build Decisions.
1. Create a decision named Practice1 in My Folder/Decisioning/Practices.

a. If necessary, click Decisions.
b. Click New Decision.
c. For Name, enter Practice1.
d. Next to Location, click Choose a location. Click My Folder  Decisioning  Practices 
OK.
e. Click Save.
2. Import variables from the comma-delimited file

D:\Workshop\ebdid\Practices\ClaimVariables.csv. Notice that the variables listed below are
created.
Variable Data Type Input Output
Airport_Code Character ✔ ✔
Claim_Num Character ✔ ✔
Incident_Date Date ✔ ✔
a. Click Variables.
b. Click Import  Comma-delimited (*.csv).
c. Click Browse.
d. Navigate to D:\Workshop\ebdid\Practices\ and select ClaimVariables.csv.
e. Click Open.
f. For Encoding, confirm that UTF-8 (Default) is selected.
g. Click Import.
h. Click Add.
i. Confirm that the Input and Output properties are enabled for both variables.
3. View the properties of the rule set named Practice_VariableMapping. Notice that it includes the
variables listed below.
Airport_Code Character ✔ ✔
Claim_Number Character ✔ ✔
Elapsed_Days Integer ✔
Incident_Date Date ✔ ✔
a. Click Rule Sets

b. Click Demo_VariableMapping to open it.
c. Click Variables and view the variable properties.
4. Close the rule set. Return to the decision and add the Practice_VariableMapping rule set to
follow the Start node.
a. Click Close to close the rule set.
b. Click Decisions to return to the decision.
c. Click Decision Flow.
d. Right-click Start and select Add below  Rule set.
e. Click My Folder  Decisioning  Practices  Practice_VariableMapping. Click OK.
5. View the decision variables. Notice that the variables Claim_Number and Elapsed_Days have
been added to the decision.
Click Variables and view the variable properties.
6. Map the rule set input variable Claim_Number to the decision input variable Claim_Num. Hint:
Click the rule set node and then click Input Variables at the right of the Properties pane for the
rule set.
a. Click Decision Flow.
b. If necessary, click the Practice_VariableMapping node.
c. Click Input Variables in the Properties pane to the right of the interface.
d. In the Maps To column for the input variable Claim_Number, select Claim_Num.
7. Save the decision and confirm removal of the Claim_Number variable from the decision. This
variable from the rule set maps to Claim_Num in the decision and is not needed as a decision
variable.
a. Click Save.
b. Click Yes in the window to confirm removal of the Claim_Number variable from the
decision.
Solution: Creating a Rule Set

In this practice, you create a rule set.
1. Create an assignment rule set named Practice_ClaimStatus in My Folder/Decisioning/Practices.

a. Click Rule Sets.
b. Click New Rule Set.
c. For Name, enter Practice_ClaimStatus.
d. For Type, confirm that Assignment is selected.
e. Next to Location, click Choose a location. Click My Folder  Decisioning  Practices 
OK.
f. Click Save.
2. Add the variables Claim_Amount and Claim_Code from the data table TSAClaims. Specify that
they are both input and output variables. The variable properties should match those shown
below.
Claim_Amount Decimal ✔ ✔
Claim_Code Character ✔ ✔
Hint: if the TSAClaims table does not appear in the list of tables on the Available tab, click
Refresh. If it still does not appear, navigate to cas-shared-default  Public and load it into
memory.
a. If necessary, click Variables.

b. Click Add variable  Data Table.
c. Select TSAClaims
d. Press the Ctrl key and click Claim_Amount and Claim_Code.
e. Click Add.
f. Click Add.
g. Confirm that the Input and Output properties are enabled for both variables.
3. Add a custom character output variable named Claim_Status. Assign a length of 8 and the initial
value PROCESS to the variable. The variable properties should match those shown below.
Variable Data Type Input Output Initial Value
Claim_Status Character ✔ PROCESS
a. Click Add variable  Custom variable.

b. For Name, enter Claim_Status.
c. For Data type, confirm that Character is selected.
d. Click Add.
e. Disable the Input property.
f. For Length, enter 8.
g. For Initial Value, enter PROCESS.
h. Click OK.
4. Define a rule named Reject missing claim codes to set Claim_Status to REJECT if the variable
Claim_Code does not have a value.
Hint: Use the expression editor to specify the MISSING function to test whether Claim_Code has
a value.
a. Click Rule Set.

b. Click Add Rule. Notice that an IF condition and an ASSIGN action are populated in the
rule set editor.
c. Rename the rule.
 Click Actions  Rename rule.
 Enter Reject missing claim codes.
d. Specify the condition for Claim_Code.
 Move the mouse pointer over the condition and click Open the expression
editor.
 Click Clear.
 Click the Functions tab.
 In the Filter box, enter missing.
 Notice that the Data Grid and Special folders are listed.
 Expand the Special folder and double-click the MISSING function to add it to the
expression.
 Notice that the argument to the function is selected.
 Click the Variables tab and double-click Claim_Code to add as the function
argument.
 Click Validate to validate the expression.
 Click Save.
e. Next to ASSIGN, select Claim_Status.
f. For the value, enter 'REJECT'.
5. Define an ELSE rule named Review high claim amounts to set Claim_Status to REVIEW if
Claim_Amount exceeds 250.
Hint: Rename the rule before changing the IF to ELSE.
a. Click Add Rule.

b. Rename the rule.
 Click Actions  Rename rule.
 Enter Review high claim amounts.
c. Click IF and select ELSE.
d. For the ELSE condition, select Claim_Amount.
e. For the operator, select > - (Is greater than).
f. For the value, enter 250.
g. Next to ASSIGN, select Claim_Status.
h. For the value, enter 'REVIEW'.
6. Import test scenarios from the file D:\Workshop\ebdid\Practices\TSAClaimScenarioInput.csv.
Hint: You use the Import Scenarios button on the Scenarios tab to import scenario tests. There
is also an Import button that appears at the top of the rule set that is used for a different
purpose.
This file contains the following records:
Claim_Code,Claim_Amount,Claim_Status_expected
,125,REJECT
PD,300,REVIEW
PP,75,PROCESS
a. Click Scoring.
b. Click Scenarios.
c. Click Import Scenarios.
d. Click in the Import from field, navigate to D:\Workshop\ebdid\Practices, and select
TSAClaimScenarioInput.csv. Click Open.
e. For Encoding, confirm that UTF-8 (Default) is selected.
f. For Scenario name prefix, enter TSAClaim.
g. For Folder Location, click Choose a location. Click My Folder  Decisioning  Practices
 OK.
h. Click Output data library. Click cas-shared-default  Public  OK.
i. Click Import.
j. Click Close in the Import Scenarios window.
k. Notice that three scenarios have been imported.
7. Run the scenarios and confirm that all three scenarios ran successfully without warnings. Close
the rule set.
a. Click Select all.
b. Click Run.
c. Confirm that the Status column contains a green check mark without a warning icon for
all three scenarios.
d. Click Close to close the rule set.
Solution: Adding a Rule Set to a Decision

In this practice, you create a decision and add a rule set.
1. Create a decision named Practice_AddRuleSet in My Folder/Decisioning/Practices.


c. For Name, enter Practice_AddRuleSet.
OK.
e. Click Save.
2. Add the rule set named Practice_ClaimStatus created in a previous practice to follow the Start
node. If you did not complete the previous practice, use the rule set named
Practice_ClaimStatusSolution1.
a. Right-click Start and select Add below  Rule set.
b. Click My Folder  Decisioning  Practices  Practice_ClaimStatus  OK.
3. On the Variables tab, confirm that Claim_Amount, Claim_Code, and Claim_Status have been
added to the decision.
a. Click Variables. Confirm that Claim_Amount, Claim_Code, and Claim_Status have been
added to the decision.
b. Click Save.
4. Define and run a basic test in the folder /Users/lynn/My Folder/Decisioning/Practices using the
data table named TSAClaims.
a. Click Scoring.
b. Click New Test.
c. Accept the default name for the test.
d. If Location is not populated with the desired path, click Choose a location. Click My
Folder  Decisioning  Practices  OK.
e. If Input table is not populated with the desired table, click Select a table and select
TSAClaims.
f. Click Run.
5. View the test results. Confirm that the values assigned to the variable Claim_Status are as
expected for the rule set.
Hint: Recall that Claim_Status should have the value REJECT when Claim_Code is missing, the
value REVIEW when Claim_Amount is greater than 250, and the value PROCESS otherwise.
a. Click Results.
b. Confirm that Claim_Status has the value REJECT when Claim_Code has a missing value.
c. Confirm that Claim_Status has the value REVIEW when Claim_Amount is greater than
250.
d. Confirm that Claim_Status has the value PROCESS for all other records.
6. Close the test results and the rule set.

a. Click Close to close the test results.
b. Click Close to close the rule set.
Solution: Creating a Lookup Table

In this practice, you create a lookup table containing claim type codes and descriptions.
1. Create a lookup table named ClaimType in My Folder/Decisioning/Practices.

a. Click Lookup tables.
b. Click New Lookup Table.
c. Specify table properties.
 For Name, enter ClaimType.
 Next to Location, click Choose a location. Click My Folder  Decisioning 
Practices  OK.
 Click Save.
2. Import key-value pairs from the file D:\Workshop\ebdid\Practices\ClaimTypeLookup.csv.

a. Click Import.
b. In the Import Lookup Table window, click Browse.
c. Navigate to D:\Workshop\ebdid\Practices and select ClaimTypeLookup.csv. Click
Open.
d. For Encoding, confirm that UTF-8 (Default) is selected.
e. Click Import.
3. Activate and close the lookup table.

a. Click Activate.
b. Click Yes.
c. Click Close to close the lookup table.
Solution: Using Lookup Table Functions in a Rule Set

In this practice, you create a new version of a rule set and add an assignment using the
LOOKUPVALUE lookup table function.
1. Open the rule set named Practice_ClaimStatus and create a new minor version.
a. Click Rule Sets.

b. Click Practice_ClaimStatus.
c. Click Versions.
d. Click New Version.
e. For Type, maintain the default selection of Minor.

f. Click Save.
2. Add a custom character output variable named Claim_Description with a length of 24. The
variable properties should match those shown below.
Variable Data Type Input Output Length
Claim_Description Character ✔ 24

b. For Name, enter Claim_Description.
c. For Data type, confirm that Character is selected.
d. Click Add.
f. For Length, enter 24.
g. Click OK.
3. Define an assignment to retrieve a value for Claim_Description from the ClaimType lookup table
using the key value Claim_Code.
If you did not create and activate the ClaimType lookup table in the previous practice, use the
ClaimTypeSolution lookup table instead.
a. Click Rule Set.

b. Click Add Rule. Notice that an IF condition and an ASSIGN action are populated in the
rule set editor.
c. Place the mouse pointer over the IF condition and click Delete the selected condition.
d. Click ASSIGN and select LOOKUPVALUE.
e. For the variable, select Claim_Description.
f. Click Select a lookup table. Navigate to My Folder  Decisioning  Practices, and
select ClaimType.
g. For the lookup key, select Claim_Code.
4. Save and close the rule set.
a. Click Save.
b. Click Close.
5. Update the Practice_AddRuleSet decision to use the new version of the rule set.
a. Click Decisions.
b. Click Practice_AddRuleSet.
c. On the Decision Flow tab, click the Practice_ClaimStatus node.
d. In the Properties pane, select 1.1 for Version.
e. Click Save to save the decision.
6. Run the previously defined test. Confirm that Claim_Description is populated with values as
expected.
a. Click Scoring.
b. Enable the check box for the previously defined test.
c. Click Run.
d. After the test finishes running, confirm that the Status column contains a green check-
mark.
e. Click Results.
f. Confirm that Claim_Description has the value Property Damage when Claim_Code has
the value PD.
g. Confirm that Claim_Description has the value Passenger Property Loss when
Claim_Code has the value PP.
7. Close the test results and the decision.

b. Click Close to close the decision.
Solution: Creating and Using a Global Variable

In this practice, you create and use a global variable.
1. Create a decimal global variable named NYC_Factor with a value of 1.15.

a. Click Global Variables.
b. Click New Global Variable.
c. For Name, enter NYC_Factor.
d. For Data Type, select Decimal.
e. For Value, enter 1.15.
f. Click Save.
2. Create a new minor version and activate the variable.

a. Click NYC_Factor to edit the variable.
b. Click Versions.
c. Click Activate.
d. Click Yes to create a new minor version.
e. Click Cancel.
3. Open the rule set named Practice_UseGlobalVariable and add the NYC_Factor variable.
a. Click Rule sets.
b. Click Practice_UseGlobalVariable to open the rule set.

c. Click Variables.
d. Click Global Variables.
e. Click Select Variables.
f. Enable the check box for NYC_Factor.
g. Click OK.
4. Modify the THEN assignment in the existing rule to use the global variable instead of a constant
1.2. The final expression should be similar to the following: Claim_Amount = Claim_Amount *
NYC_Factor.
a. Click Rule Set.
b. Move the mouse pointer over the existing THEN condition criteria and click Open the
expression editor.
c. Select the constant value 1.2 in the expression and double-click NYC_Factor in the list of
variables.
d. Click Validate and make corrections if necessary.
e. Click Save.
f. Click Save to save the rule set.
g. Click Close to close the rule set.
5. Save and close the rule set.

a. Click Save.
b. Click Close.
Solution: Configuring a Branch Node

In this practice, you configure a branch node.
1. Open the decision named Practice_Branch. Notice that the decision includes an input variable
named Claim_Amount.
b. Click Practice_Branch.
c. Click Variables.
d. Notice that the decision includes an input variable named Claim_Amount.
2. Add a Range branch node using the variable Claim_Amount following the Start node.
a. Click Decision Flow.
b. Right-click Start and select Add below  Branch.
c. For Branch type, select Range.
d. Click OK.
e. In the Properties pane, confirm that Claim_Amount is selected for the Branch variable.
3. Define a branch corresponding to values between 0 and 100.

a. In the Properties pane, click Add branch path.
b. In the Minimum field, enter 0.
c. In the Maximum field, enter 100.
4. Define a branch corresponding to values between 100 and 200.

a. Click Add branch path.
c. In the Maximum field, enter 200.
5. Define a branch corresponding to values greater than 200.

a. Click Add branch path.
6. Notice the warning icon on the branch node. Move the mouse pointer over the icon and notice
the message about overlapping values.
7. Change the label for the branch labeled 200 - No maximum to 200 or more. Hint: Click the More
link at the bottom of the branch Properties pane to change branch labels.
a. If necessary, click the Branch node.
b. In the Properties pane, click More.
c. In the Alternate Label field for the branch, enter 200 or more.
d. Click Close.
8. Save and close the decision.

a. click Save.
b. Click Close.
Solution: Configuring a Cross-Branch Link

In this practice, you configure a cross-branch link.
1. Open the decision named Practice_CrossBranchLink. Notice that the decision includes a Range
branch with four paths.
b. Click Practice_CrossBranchLink to open the decision.
2. Add the rule set named Practice_CrossBranchRuleSet to follow the existing rule set in the first
branch.
a. On the Decision Flow tab, click 0-100 and select Add  Rule set.
b. Click My Folder  Decisioning  Practices  Practice_CrossBranchRuleSet  OK.
3. Add a cross-branch link to the newly added rule set from the second branch of the node.
a. Right-click 100-200 and select Add  Cross-branch link.
b. For Target node, select Practice_CrossBranchRuleSet (1.0).
c. Click OK.

a. Click Save.
b. Click Close.
Solution: Configuring a Record Contacts Node

In this practice, you configure a Record Contacts node.
1. Open the decision named Practice_RecordContacts. Notice that the decision includes a Range
branch with four paths.
b. Click Practice_RecordContacts to open the decision.
2. Add a Record Contacts node to follow the existing rule set in the first branch. Track the variables
Claim_Number, ElapsedDays, and Item_Category.
a. On the Decision Flow tab, click Practice_CrossBranchRuleSet and select Add  Record
contacts.
b. In the Properties pane, click Select one or more variables.
c. Hold down the Ctrl key and click Claim_Number, ElapsedDays, and Item_Category.
d. Click Add.
e. Click OK.

a. Click Save.
b. Click Close.
Solution: Comparing Decision Versions

In this practice, you compare versions of a decision.
1. Open the decision named Practice_Versions.

b. Click Practice_Versions to open the decision.
2. Compare versions 1.0 and 1.1 of the decision.

a. Click Versions.
b. Click Actions  Compare object contents.
c. Maintain the default selections for base and comparison versions.
d. Click Compare.
3. View the differences on the Decision Flow tab. Notice that the two versions of the decision use
different versions of the same rule set.
a. Confirm that the Decision Flow tab is active.
b. Click Show Differences.
c. Notice that version 1.0 uses version 1.0 of the rule set Practice_AssignClaimStatus, while
version 1.1 uses version 1.1 of the same rule set.
4. View the differences on the Decision tab. Notice that both versions of the decision have a
branch node based on the variable Claim_Amount, but that version 1.0 has four output paths
while version 1.1 has three.
a. Click Decision and view the path definitions.
5. View the differences on the Variables tab. Notice that version 1.1 includes a variable that
version 1.0 does not.
a. Click Variables
b. Notice that version 1.1 includes the variable ClaimDescription while version 1.0 does
not.
6. Close the comparison results and the decision.

a. Click Close to close the comparison results.
b. Click Close to close the decision.
Lesson 3 Solutions
Solution: Creating a Data Query for Use in a Decision

In this practice, you create a data query using the SQL editor.
Reminder: Sign in to SAS using the user ID lynn and password Student1. To navigate
1. Create a data query named Practice_DataQuery using the SQL editor in My

Folder/Decisioning/Practices.
a. Click Code files.
b. Click New Code File.
c. For Name, enter Practice_DataQuery
d. For Type, select Data Query.

e. For Editor, confirm that SQL editor is selected.
f. Next to Location, click Choose a location. Click My Folder  Decisioning  Practices 
OK.
g. Click Save.
2. Modify the query properties to return scalar variables.

a. Click the Properties tab.
b. For Output Type, select Scalar.
3. Use Notepad to open the file named Practice_QueryCode.txt in the folder

D:\Workshop\ebdid\Practices. Copy the code and paste it into the Code tab following the
comment:
/* include sqlReturnInfo */.
a. Click Start  File Explorer.

b. Navigate to D:\Workshop\ebdid\Practices.
c. Double-click Practice_QueryCode.txt to open the file.
d. Hold down the Ctrl key and click A to select all the code in the file.
e. Hold down the Ctrl key and click C to copy the selected code.
f. Click the browser tab in the system tray to return to Intelligent Decisioning.
g. Click the Code tab.
h. Click in the code window below the comment in the first line.
i. Hold down the Ctrl key and click V to paste the copied code.
This code returns the columns State, Class, and ClaimsPaid from the airportdata table
when the AirportCode column matches the input variable value.
4. Validate the code and make corrections if necessary.

a. Click Validate.
b. Notice the message indicating the statement must be run. Click Run Validation.
c. Click Close to close the window showing the validation results.
d. If validation was not successful, make corrections and repeat the validation.
5. Synchronize and view the variables.

a. Click Sync Variables. Notice the message indicating that the Variables tab has been
updated.
b. Click the Variables tab.
c. Notice that the query includes the input variable Airport_Code and output variables
ClaimsPaid, Class, ReturnCode, RowCount, and State.
6. Save the code file. Create a test for the query in My Folder/Decisioning/Practices using the input
table TSAClaims. Accept the default name for the test.
a. Click Save to save the code file.
b. Click the Scoring tab.
c. Click New Test.

d. For Name, accept the default value.
e. For Location, confirm that /Users/lynn/My Folder/Decisioning/Practices is selected.
f. For Input table, click Select an input table. Click TSAClaims and click OK.
g. Click Run.
7. View the test results and confirm that the results include the variables specified in the query
SELECT clause. Close the code file.
a. When the test completes, confirm that the Status column contains a green check
indicating that the test ran successfully.
b. Click Results.
c. Notice that the output table includes the three variables specified on the SELECT clause
along with ReturnCode and RowCount.
d. Click Close to close the test results.
e. Click Close to close the code file.
Solution: Adding a Data Query to a Decision

In this practice, you create a decision and add a data query.
1. Create a decision named Practice_AddDataQuery in My Folder/Decisioning/Practices.

c. For Name, enter Practice_AddDataQuery.
OK.
e. Click Save.
2. Add the data query named Practice_DataQuery created in a previous practice to follow the Start
node. If you did not complete the previous practice, use the rule set named
Practice_DataQuerySolution.
a. Right-click Start and select Add below  Data query.
b. Click My Folder  Decisioning  Practices  Practice_DataQuery  OK.
3. Notice the settings in the Properties pane that enable you to change the output type and open
the query code file.
In the Properties pane, notice the Select output type button and the Open button under Code
file editor.

a. Click Save.
b. Click Close.
Solution: Defining DS2 Code for Use in a Decision

In this practice, you define a DS2 code file for use in a decision.
1. Create a new DS2 code file named Practice_DS2Code in My Folder/Decisioning/Practices.

b. Click New Code File.
c. For Name, enter Practice_DS2Code.
d. For Type, confirm that DS2 Code File is selected.
e. Next to Location, click Choose a location. Click My Folder  Decisioning  Practices 
OK.
f. Click Save.
2. Notice that the Code tab is populated with starter code.
The following code appears on the Code tab:
package "${PACKAGE_NAME}" /inline;

method execute();
end;
endpackage;
3. On the Variables tab, notice that there are no variables defined.

a. Click the Variables tab.
b. Notice that no variables are defined.
4. Use Notepad to open the file named Practice_DS2Code.txt in the folder

D:\Workshop\ebdid\Practices. Replace the starter code on the Code tab with the code from the
file.
a. Click Start  File Explorer.
b. Navigate to D:\Workshop\ebdid\Practices.
c. Double-click Practice_DS2Code.txt to open the file.
d. Hold down the Ctrl key and click A to select all the code in the file.
e. Hold down the Ctrl key and click C to copy the selected code.
f. Click the browser tab in the system tray to return to Intelligent Decisioning.
g. Click the Code tab.
h. Click in the code window.
i. Hold down the Ctrl key and click A to select all code in the window.
j. Hold down the Ctrl key and click V to paste the copied code.
This code determines a value for the output variable Status based on the input variables
Airport_Code, Claim_Code, Incident_Date, and Receipt_Date. The logic in this code
could be reproduced in a rule set.
5. Synchronize the variables and confirm that variables are defined on the Variables tab. Define all
variables except for Status as input variables.
a. Click Sync Variables.
b. Click the Variables tab.
c. Enable the Input property for Airport_Code, Claim_Code, Incident_Date, and
Receipt_Date.
6. Save the code file.

a. Click Save to save the code file.
b. If prompted, click Yes to sync the variables.
7. Test the decision using the data table tsaclaims. Confirm that the test runs successfully and the
output includes the output variable Status.
a. Click New Test.
b. Accept the default name for the test.
c. Confirm that /Users/lynn/My Folder/Decisioning/Practices is selected for Location.
d. For Input table, click Choose a location and select tsaclaims.
e. Click Run.
f. When the test completes, confirm that the Status column indicates that the test
completed successfully.
g. Click Results.
h. Notice the results include the column Status, which was assigned a value in the code.
i. Click Close to close the test results.
Solution: Adding a Model to a Decision

In this practice, you add a model to a decision.
1. Open the decision Practice_AddModel. Notice that it includes a data query named
GetHMEQData.
b. Click Practice_AddModel to open the decision.
c. View the decision flow.
d. On the Decision Flow tab, notice that the flow includes a data query named
GetHMEQData.
2. View the decision variables. Notice that the variable Account is both an input and output
variable, and that Loan is an input variable. All other variables are temporary. Hint: Click the
Input column heading twice to sort the variables by the presence of the Input property.
b. Click the Input column heading twice to sort the variables by the presence of the Input
property.
c. Notice that the variable Account has both the Input and Output properties enabled, and
Loan has the Input property enabled. The remaining variables have neither property
enabled.
3. In the decision flow following the data query, add the model QS_Tree1 from the model
repository named Public.
a. Click the Decision Flow tab.
b. Click the node labeled GetHMEQdata (1.0) tab.
c. Right-click and select Add  Model.
d. Navigate to Public  QS_Tree1 and select QS_Tree1. Click OK.
e. This model uses a decision tree to predict the probability of defaulting on a loan.
4. View the decision variables. Notice that the decision includes the additional input variable
Value. This is an input variable to the model that is not defined in the decision.
property.
c. Notice that Value now appears with the Input property enabled.
5. Map the model input variable Value to the decision variable Loan.
b. If necessary, click the node labeled QA_Tree1.
c. Click Input Variables.
d. Click in the Maps To column next to Value and select Loan.
6. Save the decision. Confirm the removal of the variable Value from the decision. This variable
from the model maps to the decision variable Loan and is not needed as a decision variable.
a. Click Save to save the decision.
b. Click Yes to confirm removal of the variable Value the decision.
7. Test the decision using the previously defined test. Confirm that the test runs successfully and
the output includes model results such as the variable EM_PROBABILITY.
a. Click Scoring.
b. Click the name of the existing test to open it.
c. Click Run.
d. When the test completes, confirm that the Status column indicates that the test
e. Click Results.
f. Notice the results include the column EM_PROBABILITY, which is the predicted
probability of default.
g. Click Close to close the test results.
Solution: Configuring a Data Query to Return a Data

Grid
In this practice, you configure a data query to return a data grid.
1. Open the data query named Practice_QueryReturnDataGrid. Notice that the query returns the
variables State, Class, and ClaimsPaid.
b. Click Practice_QueryReturnDataGrid.
c. On the Code tab, notice that the query code includes a SELECT clause with the variables
State, Class, and ClaimsPaid.
2. View the code variables and notice that the variables State, Class, and ClaimsPaid are defined as
output variables.
b. Notice that the variables State, Class, and ClaimsPaid are defined as output variables.
3. Modify the query properties to return a data grid instead of scalar variables.
b. For Output Type, select Data grid.
4. Synchronize the variables. View the code variables and notice that the variables State, Class,
and ClaimsPaid no longer appear and a data grid variable has been added.
a. Click the Code tab.
b. Click Sync Variables.
c. Click the Variables tab.
d. Notice that the variables State, Class, and ClaimsPaid no longer appear and the variable
dgo has been added. The variable dgo is a data grid with columns corresponding the
three columns that were specified on the SELECT clause.
5. Save the code and run the test that has been previously defined.
a. Click Save to save the code.
b. Click the Scoring tab.
c. Click the name of the existing test to open it.
d. Click Run.
6. View the test results. Confirm that the output data grid contains the columns State, Class, and
ClaimsPaid.
Note: RowCount has a value of 0 or 1 for each row in the input table. For this test data,
at most one row matches the query for each row in the input data. Therefore, the
corresponding data grids all have at most a single row. However, it is possible for a data
grid to include multiple rows.
a. When the test completes, confirm that the Status column contains a green check
indicating that the test ran successfully.
b. Click Results.
c. Notice that the output table includes the data grid along with ReturnCode and
RowCount.
d. Click in the data grid column for the first row to view the contents of the data grid.
Notice that data grid includes columns correspond to the three variables specified on
the SELECT clause.
e. Click Close to close the data grid.
Solution: Using Data Grid Functions

In this practice, you use data grid functions to calculate the maximum value of a data grid
column and create a sorted copy of a data grid.
1. Open the decision Practice_DataGridFunctions. Notice that it includes a data query named
OrderQuery that returns a data grid.
b. Click Practice_DataGridFunctions to open the decision.
 On the Decision Flow tab, notice that the flow includes a data query named
OrderQuery.
 Click the OrderQuery node and notice that the data query is configured to
return a data grid.
2. View the properties of the decision variables to identify the columns in the data grid.
b. Click OrderQuery_out to view the properties of the data grid.
c. Notice that the data grid includes the columns Profit, Total_Retail_Price, Employee_ID,
Order_ID, Order_Date and Quantity.
d. Click Cancel.
3. Create an assignment rule set named Practice_DGF in the folder My

a. Click Rule sets.
c. For Name, enter Practice_DGF.
e. For Location, click Choose a location and navigate to My Folder  Decisioning 

Practices. Click OK.
f. Click Save.
4. In the rule set, add the OrderQuery_out data grid variable from the Practice_DataGridFunctions
decision and enable the Input property. The variable properties should match those shown
below.
OrderQuery_out Data grid ✔ ✔
a. Click Add variable  Decision.

b. Navigate to My Folder  Decisioning  Practices.
c. Click Practice_DataGridFunctions.
d. Click OK.
e. Click OrderQuery_out.
f. Click Add.
g. Click Add.
h. Enable the Input Property.
5. In the rule set, duplicate the OrderQuery_out data grid variable to create a variable named
OrderQuery_Desc_Profit. The variable properties should match those shown below.
OrderQuery_Desc_Profit Data grid ✔
a. Enable the check box next to OrderQuery_out.

b. Click Actions  Duplicate.
c. For Name, enter OrderQuery_Desc_Profit.
d. Click Duplicate.
6. In the rule set, add a custom decimal output variable named MaximumProfit. The variable
properties should match those shown below.
MaximumProfit Decimal ✔

b. For Name, enter MaximumProfit.
c. For Data type, select Decimal.
d. Click Add.
f. Click OK.
7. Create an assignment to populate the MaximumProfit variable with the maximum value of the
Profit column of the OrderQuery_out data grid. Hint: Use the DATAGRID_MAX function.
a. Click Rule Set.
b. Click Add Assignment.
c. Next to ASSIGN, select MaximumProfit.
d. Click Open the expression editor.
e. Click Functions.
f. Expand the Data Grid folder.
g. Double-click DATAGRID_MAX to add it to the expression.
h. Click Variables.
i. Double-click OrderQuery_out to add it as the first argument to the function.
j. Select colName in the second argument and enter Profit. The expression should appear
as shown below:
MaximumProfit = DATAGRID_MAX(OrderQuery_out,'Profit')
k. Click Validate to validate the expression. Make corrections if necessary.

l. Click Save.
8. Create an assignment to populate the OrderQuery_Desc_Profit data grid with the values from
the OrderQuery_out data grid in descending order of profit. Hint: Use the DATAGRID_SORT
function.
a. Click Add  Add assignment.
b. Hover over the newly added assignment and click Open the expression editor.
c. Click Clear to clear the expression.
d. Click Functions.
e. Expand the Data Grid folder.
f. Double-click DATAGRID_SORT to add it to the expression.
g. Click Variables.
h. Double-click OrderQuery_out to add it as the first argument to the function.
i. Select sortCol in the next argument and enter Profit.
j. Select sortOrder in the next argument and enter d.
k. Select tgtGrid in the final argument. Click Variables and double-click
OrderQuery_Desc_Profit. The expression should appear as shown below:
DATAGRID_SORT(OrderQuery_out,'Profit','d',OrderQuery_Desc_Profit)
l. Click Validate to validate the expression. Make corrections if necessary.

m. Click Save.
9. Save and close the rule set. Add the rule set to the Practice_DataGridFunctions decision
following the data query. Save the decision.
a. Click Save.
b. Click Close.
c. Click Decisions to return to the Practice_DataGridFunctions decision.
d. Click Decision Flow.
e. Right-click the OrderQuery node and select Add  Rule set.
f. Navigate to My Folder  Decisioning  Practices.
g. Click Practice_DGF.
h. Click OK.
i. Click Save to save the decision.
10. Test the decision using the previously defined test named Practice_DataGridFunctions_Test_1.
Confirm that the MaximumProfit variable has values and that the OrderQuery_Desc_Profit data
grid is sorted by descending profit.
a. Click Scoring.
b. Click the test named Practice_DataGridFunctions_Test_1 to open it.
c. Click Run.
e. Click Results.
f. Confirm that the MaximumProfit column is populated with values.
g. Click OrderQuery_Desc_Profit to open the data grid for the first row in the output table.
Confirm that the data grid rows are sorted by descending profit. Click Close.
h. Click Close to close the test results.
11. Close the decision.

Click Close.
Solution: Scoring Rows in a Data Grid

In this practice, you create a rule set to score rows in a data grid to calculate a
commission for products with a price of more than 200.
1. Open the decision Practice_DataGridScoreRows. In the data grid named

OrderQuery_Desc_Profit, define a decimal column named Commission. Save the decision.
b. Click Practice_DataGridScoreRows to open the decision.
c. Click the Variables tab.
d. Click OrderQuery_Desc_Profit to edit the data grid.
e. Notice that the data grid includes the columns Profit, Total_Retail_Price, Employee_ID,
Order_ID, Order_Date and Quantity.
f. Click Edit.
g. Confirm that Add a new column is selected.
h. Enter the name Commission.
i. Select Decimal.
j. Click Add.
k. Click OK to close the Edit Columns window.
l. Click OK to close the Edit Variable window.
m. Click Save to save the decision.
2. Create an assignment rule set named Practice_DGS in the folder My

a. Click Rule sets.
c. For Name, enter Practice_DGS.
f. Click Save.
3. In the rule set, add a custom decimal input variable named Price. Add a custom decimal output
variable named Commission. The variable properties should match those shown below.
Price Decimal ✔
Commission Decimal ✔
a. Define the Price variable.

 Click Add variable  Custom variable.
 For Name, enter Price.
 For Data type, select Decimal.
 Click Add.
 Disable the Output property.
 Keep the Add Variables window open.
b. Define the Commission variable.
 For Name, enter Commission.
 For Data type, maintain the selection Decimal.
 Click Add.
 Disable the Input property.
 Click OK.
4. Define a rule to calculate the value of Commission. If Price is greater than 200, then
Commission is 5% of Price. Otherwise, Commission is 0.
a. Click Rule Set.
b. Define the IF rule.

 Click Add Rule.
 Next to IF, select Price.
 For the operator, select Is greater than.
 Enter the value 200.
 Next to THEN, confirm that ASSIGN is selected.
 For the variable, select Commission.
 Enter the expression Price*0.05.
c. Define the ELSE rule.
 Click Add  ELSE rule.
 Next to THEN, confirm that ASSIGN is selected.
 For the variable, select Commission.
 Enter the value 0.
5. Save and close the rule set. Add the rule set to the Practice_DataGridScoreRows decision
following the rule set named Practice_DGFSolution.
a. Click Save.
b. Click Close.
c. Click Decisions to return to the DataGridDemo decision.
d. Click Decision Flow.
e. Right-click the Practice_DGFSolution node and select Add  Rule set.
f. Navigate to My Folder  Decisioning  Practices.
g. Click Practice_DGS.
h. Click OK.
6. Configure the rule set to score rows in the OrderQuery_Desc_Profit data grid. Map the input
variable named Price to a data grid column named Total_Retail_Price. Map the scalar output
variable named Commission to a data grid column of the same name. Hint: You perform these
steps after clicking Input Variables in the Properties pane for the rule set in the decision flow.
a. Click Input variables.
b. Enable the Score rows in this data grid option.
c. Confirm that the data grid named OrderQuery_Desc_Profit is selected.
d. Configure variable mapping.
 Click in the Maps To column next to Price and select the data grid column
Total_Retail_Price.
Note: If the variable had been named Total_Retail_Price, this mapping

would have been configured by default.
 Click Output variables. Confirm that the output variable Commission maps to a
data grid column of the same name.
7. Save the decision. Confirm the removal of any variables no longer referenced by objects in the
decision.
a. Click Save.
b. Click Yes.
8. Test the decision using the previously defined test named Practice_DataGridScoreRows_Test_1.
Confirm that the Commission column of the OrderQuery_Desc_Profit data grid has a positive
value for items with Total_Retail_Price greater than 200, and 0 otherwise.
a. Click Scoring.
b. Click the test named Practice_DataGridScoreRows_Test_1 to open it.
c. Click Run.
e. Click Results.
f. Click OrderQuery_Desc_Profit to open the data grid for the first row in the output table.
Confirm that the Commission column has a positive value for items with
Total_Retail_Price greater than 200, and 0 otherwise. Click Close.
g. Click Close to close the test results.

Click Close.
Lesson 4 Solutions
Solution: Creating and Locking a Treatment

In this practice, you create and lock a treatment.
to SAS Intelligent Decisioning, click Show list of applications, Build Decisions.
1. Create a treatment named PracticeTreatment in My Folder/Decisioning/Practices.

a. Click Treatments.
b. Click New Treatment.
c. For Name, enter PracticeTreatment
OK.
e. Click Save.
2. Add all attributes from the treatment named StandardTreatment.

a. Click Add Attribute  Treatment.
b. Navigate to My Folder  Decisioning  Practices and select StandardTreatment. Click
OK.
c. Click Add all.
d. Click Add.
e. Notice that the attributes Category and Priority are added.
3. Change the value of the Priority attribute to 5.

a. Click Priority
b. In the Value field, enter 5.

c. Click OK.
4. Add a dynamic URL custom attribute named LinkURL.

a. Click Add Attribute  Custom attribute.
b. For Name, enter LinkURL.
c. For Data type, select URL and Dynamic.
d. Click Add.
e. Click OK.
5. Add a fixed character custom attribute named Description with the value Practice Treatment.
a. Click Add Attribute  Custom attribute.
b. For Name, enter Description.
c. For Data type, select Character and Fixed.
d. For Value, enter Practice Treatment.
e. Click Add.
f. Click OK.
6. Specify that the treatment is effective until 12 pm on the last day of the upcoming month.
b. Next to End date, click Select a date and time.
c. Select the last day of the upcoming month.
d. For time, select 12, 00, 00, PM.
e. Click OK.
7. Create a locked version for use in a treatment group. Close the treatment.
a. Click the Versions tab.
b. Click New Version.
c. Maintain the default selection of Minor.
d. Click Save.
e. Notice that version 1.1 is created and version 1.0 is locked.
f. Click Close.
Solution: Creating a Treatment Eligibility Rule

In this practice, you create a filtering rule set and specify it as the eligibility rule for a
treatment.
1. Create a filtering rule set named Practice_EligibilityRule in My Folder/Decisioning/Practices.

a. Click Rule sets.
c. For Name, enter Practice_EligibilityRule.
d. For Type, select Filtering.
f. Click Save.
2. Add the variable Airport_Code from the data table TSAClaims.

a. Click Add variable  Data table.
b. If necessary, click Select data table, select TSAClaims and click OK.
c. Click Airport_Code.
d. Click Add.
e. Click Add.
3. Define a rule that is true when Airport_Code does not have the value JFK, EWR, or LGA.
a. Click the Rule Set tab.
b. Click Add Rule.
c. Notice that the variable Airport_Code is populated in the IF condition.
d. For the operator, select NOT IN.
e. For the value, enter 'JFK','EWR','LGA'.
4. Name the rule Not an NYC airport.

a. Click Actions  Rename rule.
b. Enter Not an NYC airport.
c. Click Rename.
5. Create a locked version of the rule set for use in a treatment. Close the rule set.
a. Click the Versions tab.
b. Click New Version.
c. Maintain the default selection of Minor.
d. Click Save.
f. Click Close.
6. Specify that the newly created rule set is the eligibility rule for the treatment named
Practice_Treatment. Save and close the treatment.
a. Click Treatments.
b. Click Practice_Treatment to open it.
c. Click the Eligibility Rule Set tab.
d. Click Add Rule Set.
e. Navigate to My Folder  Decisioning  Practices and select Practice_EligibilityRule.
f. Click OK.
g. Click Save.
h. Click Close.
Solution: Creating a Treatment Group

In this practice, you create a treatment group and activate it.
1. Create a treatment group named Practice_TreatmentGroup in My Folder/Decisioning/Practices.

a. Click Treatment Groups.
b. Click New Treatment Group.
c. For Name, enter Practice_TreatmentGroup.
d. For Location, click Choose a location and navigate to My Folder  Decisioning 
e. Click Save.
2. Add the treatments AutoFinanceOffer, DebtConOffer, and HomeImpOffer.

a. Click Add Treatments.
b. Enable the check boxes for the following treatments:
 AutoFinanceOffer
 DebtConOffer
 HomeImpOffer
c. Click OK.
3. View the properties of the AutoFinanceOffer treatment. Identify the version number, attributes,
and eligibility rule for the locked version. Note: The remaining two treatments have similar
settings.
a. Click AutoFinanceOffer.
b. Click Versions and notice that version 1.0 is locked.
c. Click in the row for version 1.0 and click Set Version. Notice that version 1.0 is now the
displayed version.
d. Click Attributes. Notice that the treatment has fixed attributes Description and URL and
the dynamic attribute Value.
e. Click Eligibility Rule Set and notice that the eligibility rule is named Purpose = Auto
Finance and includes the condition If PURPOSE = 'AF'.
f. Click Close to close the treatment.
4. In the treatment group, select version 1.0 of each treatment.

a. Click Treatment Groups to return to the treatment group.
b. In the Version column, select 1.0 for each treatment.
5. Click Set Attributes and verify the attribute settings for the treatments. Important: You must
click each treatment and then click OK (not Cancel) to confirm the dynamic treatment attribute
settings. If you do not do so, the dynamic attribute will not appear as an input variable for the
treatment group.
a. Click Set Attributes.
b. Notice that the AutoFinanceOffer treatment is displayed. Notice that as seen previously
the treatment has fixed attributes Description and URL and the dynamic attribute
Value.
c. Click each of the two remaining treatments and notice that they all share the same
three attributes.
d. Click OK.
6. Identify the eligibility variables.

a. Click Eligibility Variables.
b. Notice that there is one eligibility variable named PURPOSE. It applies to each of the
three treatments.
7. Save the treatment group. Create a new minor version and activate version 1.0 of the treatment
group. Close the treatment group.
a. Click Save.
b. Click Versions.
c. Click New Version.
d. Accept the default type of Minor. Click Save.
f. Click in the row for version 1.0.
g. Click Activate. Click Yes to confirm.
h. Click Close.
Solution: Adding a Treatment Group to a Decision

In this practice, you add a treatment group to a decision.
1. Open the decision named Practice_AddTreatmentGroup. Notice that the decision includes a
data query, a model, and a range branch.
b. Click Practice_AddTreatmentGroup.
2. View the decision variables. Notice that the decision includes the input variables Account, Loan
and Purpose. Hint: Click the Input column heading twice to sort the variables by the presence of
the Input property.
property.
c. Notice that the variables Account, Loan and Purpose have the Input property enabled.
3. Add the Practice_TreatmentGroup treatment group to the path of the decision flow labeled Low
default probability.
b. Click the Range node and select Add to branch path  Low default probability 
Treatment group.
c. Navigate to My Folder  Decisioning  Practices and select Practice_TreatmentGroup.
d. Click OK.
4. View the input variables to the treatment group. What input variables appear, and why?
a. In the Properties pane, click Input variables.
b. Notice that the treatment group includes the input variables Purpose, and Value.
 Purpose is an input to the treatment group because it used in treatment
eligibility rules.
 Value is an input to the treatment group because it is a dynamic treatment
attribute.
5. Confirm that the treatment group input variable Purpose is mapped to a decision variable of the
same name. Map the treatment group input variable Value to the decision variable Loan.
a. In the Maps To column for Purpose, confirm that the decision flow variable Purpose is
selected.
b. Click in the Maps To column for Value and select More  Loan. Click OK.
6. Save the decision and confirm removal of the variable Value. This variable has been mapped to
the decision variable Loan and is no longer needed at the decision level.
a. Click Save.
b. Click Yes.
7. View the output variables for the treatment group. What output variable appears?
a. If necessary, click the node for the treatment group.
b. In the Properties pane, click Output Variables.
c. Notice the treatment outcome maps to a data grid named
Practice_TreatmentGroup_out.
d. Notice that the columns of the data grid include standard treatment properties such as
treatment_definition_name and treatment attributes such as Value.
Solution: Using an Assignment Rule Set for Treatment

Arbitration
In this practice, you arbitrate treatments in a decision.
1. Open the decision named Practice_TreatmentArbitration. Notice that the decision includes a
treatment group and a rule set at the end of one of the paths. The rule set performs arbitration
of the treatments in the group.

b. Click Practice_TreatmentArbitration.
2. Notice that the treatment group output data grid is named Practice_AccountOffers_out and
that it contains a column named Priority. Hint: Click the treatment group node in the decision
flow and then click Output variables in the Properties pane to view the output data grid and its
columns.
a. Click the treatment group node named Practice_AccountOffers.

b. In the Properties pane, click Output variables.
c. Notice that the output data grid is named Practice_AccountOffers_out.
d. Notice that the output data grid includes the column Priority.
3. Open the treatment group named Practice_AccountOffers. Notice that Priority is a fixed
numeric attribute of each of the three treatments in the group. Notice also that the treatment
group includes eligibility rules. Close the treatment group.
a. In the Properties pane, click Node properties.

b. Click Open to open the treatment group editor.
c. Notice that the treatment group includes three treatments.
d. Click Set Attributes.
e. In the Set Attributes window, click each treatment. Notice that Priority is fixed and has a
different value for each treatment.
f. Click Cancel to close the Set Attributes window.
g. Click the Eligibility Variables tab and notice that the treatment group includes eligibility
variables.
h. Click Close to close the treatment group.
4. Open the rule set named Practice_Arbitration. Identify the input and output variables for the
rule set and the assignment that it uses. Close the rule set.
a. Click the rule set node named Practice_Arbitration.
b. In the Properties pane, click Open.

c. In the rule set, click the Variables tab. Notice that the rule set includes an input data
grid variable named Practice_AccountOffers_out and an output data grid variable
named SortedOffers.
d. Click the Rule Set tab. Notice that the rule set includes an assignment that uses the
DATAGRID_SORT function to assign the rows of the Practice_AccountOffers_out data
grid in ascending sorted value of Priority to the SortedOffers data grid.
e. Click Close to close the rule set.
5. Run the test that was previously defined. Compare the Practice_AccountOffers_out and
SortedOffers data grid variables for a row in the test data where they have values. Note: The
data grids have values only for the rows in the test data that followed the path in the decision
where the treatment group and arbitration rule set were added.
a. Click the Scoring tab.

b. Click the name of the existing test to open it.
c. Click Run.
e. Click Results.
f. Click in the row for Practice_AccountOffers_out for a row where it is populated.
g. Note the number of rows in the data grid. This number corresponds to the number of
treatments for which this row of data met the eligibility rules.
h. Scroll to the right. Take note of the values of Priority and whether they appear in sorted
order. Note: Because Priority is a fixed treatment attribute, each treatment always has
the same priority value. But because the treatments have eligibility rules, the list of
treatments sorted by priority can be different for each row of test data.
i. Click in the same row for SortedOffers. Scroll to the right and confirm that the rows are
sorted in order of Priority.
Lesson 5 Solutions
Solution: Basic Testing of a Decision

In this practice, you test a decision using rule-fired analysis, decision path tracking, and
advanced options.
1. Open the decision Practice_BasicTesting. Notice that the decision includes a data query, a rule
set, and a range branch with a rule set on each of three output paths.

b. Click Practice_BasicTesting to open the decision.
2. View the properties of the rule set named Practice_AssignQueue. Notice that the rule set
populates the output variable Queue based on the value of the input variable MortDue as
depicted below.
MortDue Value Queue Value
>120,000 A
>60,000 and <=120,000 B
>0 and <=60,000 C
a. On the Decision Flow tab, click the node named Practice_AssignQueue.

c. Click the Variables tab. Notice that MortDue is an input variable and Queue is an output
variable.
d. Click the Rule Set tab. Notice that the rule set assigns one of three values to the variable
Queue based on the value of the variable MortDue.
3. Return to the decision and run the test that has been previously defined.
c. Click the Scoring tab.
d. Enable the check box next to the test and click Run.
4. Open the test results and run rule-fired analysis. Confirm that the results match what you would
expect for the Practice_AssignQueue rule set.
a. Click Results.
b. Click Rule-Fired Analysis.
c. Click Run Rule-Fired Analysis.
d. For the first few rows, notice the value of MortDue and click the link in the Rules Fired
Count column to view details of the rule that fired.
Note: The value of the column Rule Order listed in the Rule Fired Count
window reflects the static ordering in the diagram and not necessarily the order
that the rule fired at execution time.
5. Return to the decision flow. View the variable and ranges specified in the branch node.
b. Click the Decision Flow tab.
c. Click the node labeled Range (DEBTINC).

d. Notice that the branch variable is DebtInc.
e. In the Properties pane, click More.
f. Notice that there are two branches defined, along with an Other branch.
Alternate Label Minimum Maximum
Low No minimum 36
Caution 36 43
g. Click Close to close the More Branching Properties window.
6. View the properties of the rule set named Practice_LowDebtInc. Notice that it includes an
assignment for the variable Message. Note: This simple rule set is included only to produce path
tracking results, and the rule sets on the other branch paths are similar.
a. On the Decision Flow tab, click the node named Practice_LowDebtInc.
c. Notice that the rule set assigns the value This is the Low branch to the variable Message.
7. Return to the decision. Open the previously defined test and run path tracking. View the plot
and the node counts.
c. Click the Scoring tab.
d. Click Results to open the results for the test that has been previously run.
e. Click Decision Path Tracking.
f. Click Run Path Tracking.
g. In the plot, notice that 26 rows in the test data followed the Low path, 12 followed the
Caution path, and three followed the Other path.
h. Click Node Count. Notice that the results list the count of rows that passed through
each node in the decision flow.
8. Duplicate the existing test. Enable the option Record variable values by node. Run the test and
open the test results.
a. Enable the check box next to the existing test.
c. Accept the default name for the duplicate.
d. Click Duplicate.
e. Click the name of the duplicated test to open it.
f. Click Advanced.
g. Enable Record variable values by node
h. Click Run.
i. After the test runs, confirm that the Status column displays a green check to indicate
that the test ran successfully.
j. Click Results.
9. Click nodeTraceDataGrid to see variable changes by node. Notice that a row appears for the
data query node and for each of two rule sets. Notice that most of the variable values do not
change across the nodes, but that values for the variables Queue and Message are assigned in
the rule sets. Close the data grid.
a. Click a row in the column for the nodeTraceDataGrid to open it.
b. Scroll right to view the variable values.
c. Click Close.

a. Click Close.
b. Click Close.
Solution: Scenario Testing

In this practice, you perform scenario testing of a rule set.
1. Open the rule set named Practice_ScenarioTesting. Notice that the rule set assigns a value to
the variable Category based on the variable Price as shown below.
Price Category
<50 Low
<=50 and <100 Mid
>=100 High
a. Click Rule sets.

b. Click Practice_ScenarioTesting to open the rule set.
c. Notice that the rule set assigns one of three values to the variable Category based on
Price.
2. Create a scenario test with cas-shared-default, Public as the output data library.
a. Click Scoring.
b. Click Scenarios.
c. Click New Test.
d. Notice that Output table location is required and does not have a value. Click Output
data library and select cas-shared-default  Public. Click OK.
3. Specify that the input variable Price has the value 150. Include the output variable Category
with the value High as the expected output. Run the test and confirm that it runs successfully
and that the expected and actual values of Category match.
a. For Price, enter 150 in the Value field.
b. For Category, enable the Include check box and enter High in the Expected Output field.
c. Click Run.
d. When the test completes, confirm that the Status column displays a green check mark
indicating that the scenario test completed successfully.
e. Click Results.
f. Notice that the expected and actual values for Category match.
g. Close the test results.
4. Duplicate the scenario test. Change the expected output value for Category to high (all
lowercase). Run the test and confirm that it runs with warnings and that the expected and
actual values of Category do not match.
a. Enable the check box next to the existing test.
c. Accept the default name for the test and click Duplicate.
d. Click the name of the duplicated test to open it.
e. For Category, enter high (all lowercase) in the Expected Output field.
f. Click Run.
g. When the test completes, notice the warning icon over the green check-mark in the
Status column. This icon indicates that the scenario test completed with warnings.
h. Click Results.
i. Notice the message There are some actual output values that do not match the expected
output values.
j. Click Show Differences. Notice that the expected and actual values for Category do not
match.
5. Close the test results and the rule set.

a. Click Close.
b. Click Close.
Solution: Publishing and Validating a Published

Decision
In this practice, you publish and validate a decision.
1. Open the rule set named Practice_PublishValidate.

b. Click Practice_PublishValidate to open the decision.
2. Lock the decision and publish to the maslocal destination. Confirm that publication is successful.
a. Click Actions  Publish and Lock.
b. Notice that you can change the value for Published name.
c. Notice that the destination SAS Micro Analytic Service (maslocal) is selected by default.
d. Click Publish and Lock.
e. When the process completes, confirm that the Status column indicates that the decision
was published successfully.
f. Click Close.
3. Run the publishing validation test using the input table HMEQApps. Confirm that the test runs
successfully.
a. Click the Scoring tab.
b. Click the Publishing Validation tab.
c. Notice that a publishing validation test has been created. Place your mouse pointer over
the test name and notice that it is the name of the decision with a datetime stamp
added.
d. Click the name of the test to open it.
e. Under Input table, click Select a table.
f. Click HMEQApps. Click OK.
g. Click Run.
h. When the test completes, confirm that the status column indicates that the test ran
successfully.
4. Open the test results. Notice that the publishing validation test creates an output table similar
to basic testing results. What optional analyses do not appear in the publishing validation test
for the maslocal destination that did appear in basic tests on CAS?
a. Click Results.
b. Under Test Results, notice that Rule-fired Analysis and Decision Path Tracking do not
appear.

a. Click Close.
b. Click Close.

LWBDID3 001 SolutionsToPractices

Uploaded by

Copyright:

Available Formats

LWBDID3 001 SolutionsToPractices

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LWBDID3 001 SolutionsToPractices

Uploaded by

Copyright:

Available Formats

Essential Functions of SAS Intelligent Decisioning: Solutions to Practices

Solution: Working in SAS Intelligent Decisioning

5. Access the Help Center.

6. Sign out of SAS Intelligent Decisioning.

Solution: Loading Tables into CAS

2. Repeat the previous steps to load the tables Investments.sashdat, TSAClaims.sashdat,

Solution: Managing and Mapping Variables in a

1. Create a decision named Practice1 in My Folder/Decisioning/Practices.

2. Import variables from the comma-delimited file

Variable Data Type Input Output

Variable Data Type Input Output

a. Click Rule Sets

Solution: Creating a Rule Set

1. Create an assignment rule set named Practice_ClaimStatus in My Folder/Decisioning/Practices.

Variable Data Type Input Output

a. If necessary, click Variables.

Variable Data Type Input Output Initial Value

Claim_Status Character ✔ PROCESS

a. Click Add variable  Custom variable.

a. Click Rule Set.

Hint: Rename the rule before changing the IF to ELSE.

a. Click Add Rule.

6. Import test scenarios from the file D:\Workshop\ebdid\Practices\TSAClaimScenarioInput.csv.

This file contains the following records:

Solution: Adding a Rule Set to a Decision

1. Create a decision named Practice_AddRuleSet in My Folder/Decisioning/Practices.

b. Click New Decision.

6. Close the test results and the rule set.

Solution: Creating a Lookup Table

1. Create a lookup table named ClaimType in My Folder/Decisioning/Practices.

2. Import key-value pairs from the file D:\Workshop\ebdid\Practices\ClaimTypeLookup.csv.

3. Activate and close the lookup table.

Solution: Using Lookup Table Functions in a Rule Set

a. Click Rule Sets.

e. For Type, maintain the default selection of Minor.

Variable Data Type Input Output Length

a. Click Add variable  Custom variable.

a. Click Rule Set.

4. Save and close the rule set.

e. Click Save to save the decision.

7. Close the test results and the decision.

a. Click Close to close the test results.

Solution: Creating and Using a Global Variable

1. Create a decimal global variable named NYC_Factor with a value of 1.15.

2. Create a new minor version and activate the variable.

b. Click Practice_UseGlobalVariable to open the rule set.

5. Save and close the rule set.

Solution: Configuring a Branch Node

3. Define a branch corresponding to values between 0 and 100.

4. Define a branch corresponding to values between 100 and 200.

5. Define a branch corresponding to values greater than 200.

8. Save and close the decision.

Solution: Configuring a Cross-Branch Link

4. Save and close the decision.

Solution: Configuring a Record Contacts Node

3. Save and close the decision.

Solution: Comparing Decision Versions

1. Open the decision named Practice_Versions.

b. Click Practice_Versions to open the decision.