Update the Job to Use the Latest Champion Model

This exercise demonstrates the steps to update the SAS Data Integration Studio job after you change the champion model in SAS Model Manager. After you publish the Loan project from SAS Model Manager, SAS Data Integration Studio recognizes a new mining results object.
To update the job, follow these steps:
  1. Close and reopen Tutorial11.
    1. Click the Tutorial11 window and select Filethen selectClose.
    2. Click the Folders tab and expand My Folder. Double-click Tutorial11 to reopen the job. When the job reopens, it uses the new score code from the Mining Result object and regenerates the code that is associated with the job.
  2. In the Tutorial11 diagram, right-click the Model Scoring node and select Properties. The Model Scoring Properties window appears. Click the Models tab. The Loan mining result is highlighted. The Algorithm box shows that the model is a Regression model.
  3. Click the Model Attributes tab. A message box might appear that warns of potential table changes if you change the mining result. Click Yes.
  4. Click View Source Code. Scroll to the top of the window. Compare the text in the comment tags to the Reg 1 model code in SAS Model Manager. They are the same, as shown in the displays below.
    To view the model code in SAS Model Manager, follow these steps:
    1. Log on to SAS Model Manager and expand the following Project Tree nodes:
      • Tutorial3 folder
      • Loan project
      • 2012 version
      • Models folder
      • Reg 1 model
    2. In the Reg 1 model, select score.sas. The model code appears in the Content view.
    Here is the Reg 1 score code in SAS Model Manager.
    Reg 1 Model Code in SAS Model Manager
    Here is the Reg 1 score code in SAS Data Integration Studio.
    Reg 1 Model Code in SAS Data Integration Studio
  5. Click Cancel to close the View Source Code window.
  6. Identify the variables to be used in the transform output. Click the Target Table Columns tab. If score and customer_id are not in the Selected list, follow these steps:
    1. From the Available list, select score and click Right Arrow Button to move the score variable to the Selected list.
    2. From the Available list, expand SCORE_INPUT, select customer_id, and click Right Arrow Button to move the customer_id variable to the Selected list.
  7. Click the Mappings tab. Right-click the space between the two lists of variables and select Map All. Here is the Mappings tab:
    Remapping of variables
    Click OK.
  8. Double-click the Table Loader node and click the Mappings tab.
  9. Right-click the space between the two lists of variables and select Map All. Click OK.
  10. On the Diagram tab, and click Run. The Tutorial11 job runs.
  11. To view the output, right-click the SCORE_OUTPUT node and select Open. Here is a partial view of the output:
    Partial Output from running the Reg 1 model.
Copyright © SAS Institute Inc. All rights reserved.