Organize the Model Hierarchy

Create Folders

Create a top-level folder for the quick start tutorials:
  1. Select Modelsthen selectProjects.
  2. Click Create a new item and select New Top-Level Folder. The New Folder window appears.
    New Folder
  3. Enter Tutorials for the name of the folder.
  4. (Optional) Enter a description for the folder.
  5. Click Save.
Because multiple users might want to perform the tasks in the tutorial, it is recommended that each user create their own folder in the Tutorials folder. To create a new folder:
  1. Select Tutorials, click Create a new item, and select New Folder . The New Folder window appears.
    New Folder
  2. Enter a name for the folder, such as myUserID. The examples in this tutorial use the ID sasdemo.
  3. (Optional) Enter a description for the folder.
  4. Click Save.

Create a Project

To create a project:
  1. Select a folder to store the new project (for example, myUserID).
  2. Click New and select New Project. The New Project window appears.
    New Project
  3. Enter HMEQ for the name of the project.
    The initial version is displayed and reflects the level for sequential versions.
  4. Select Classification for the model function.
    Note: The model function (Classification, Prediction, Segmentation, or Analytical) indicates the type of models that should be imported into the project.
  5. Click Save.

Import Project Variables

To import project variables:
  1. Select Variablesthen selectInput and click Import variables from a table. The Select Data Source window appears.
    Select Data Source window
  2. Select HMEQ_PROJECT_INPUT as the data source from the QSTutorial library. Click OK.
  3. Select the Output tab and click Import variables from a table.
  4. Select the HMEQ_PROJECT_OUTPUT as the data source from the QSTutorial library and click OK.
  5. Click Save to make the changes effective for other pages.
  6. Click Yes in the warning message since you have not set a champion or challenger yet.

See Also

Set the Project Properties

To define the properties that SAS Model Manager uses to create reports, score, publish, and monitor models:
  1. Select Propertiesthen selectSpecific.
  2. Select the default data tables from the QSTutorial library and specify the model variables for the project:
    Default test table
    select HMEQ_TEST.
    Default scoring input table
    select HMEQ_SCORE_INPUT.
    Default scoring output table
    select HMEQ_SCORE_OUTPUT.
    Default train table
    select HMEQ_TRAIN.
    Training target variable
    enter BAD.
    Target event value
    enter 1.
    Class target level
    select Binary.
    Output event probability variable
    select score.
  3. Click Save.
Here is an example of the HMEQ project-specific properties:
Project Specific Properties