Cross-sectional Data Models Task: Zero-inflated Poisson Model

• Example: Zero-inflated Poisson Model
• Zero-inflated Poisson Model: Assigning Data to Roles
• Zero-inflated Poisson Model: Setting the Model Options
• Setting the Options
• Zero-inflated Poisson Model: Creating Output Data

Example: Zero-inflated Poisson Model

To create this example:

1. Create the Work.Long97Data data set. For more information, see Long97Data Data Set.
2. On a Program tab, run this code to create a connection to Cloud Analytic Services (CAS) and to start a CAS session:

   options cashost="<cas-server-name>" casport=<port-number>;
   cas;

   Note: You can also use the Connect to CAS task or the Create CAS Connection snippet in SAS Studio.
3. Run this code to load the data into a CAS table:
   libname mycas cas; proc casutil; load data=work.long97data casout="long97data"; run;

   libname mycas cas;
   
   proc casutil;
      load data=work.long97data casout="long97data";
   run;

4. In the Tasks section, expand the Viya FoundationEconometrics folder, and then double-click Cross-sectional Data Models. The user interface for the Cross-sectional Data Models task opens.
5. On the Data tab, select MYCAS.LONG97DATA.
   TipIf the data source is not available from the drop-down list, click . In the Choose a Table window, expand the library that contains the data source that you want to use. Select the data source for the example and click OK. The selected data source now appears in the drop-down list.
6. Assign columns to these roles:

   Role Assignments for Zero-Inflated Poisson

   Role	Column Name
   Dependent variable	art
   Continuous variables	ment phd
   Categorical variable	kid5

7. On the Model tab, select Zero-inflated Poisson as the model type.
8. To run the task, click Run .

Here are the results:

Zero-inflated Poisson Model: Assigning Data to Roles

To perform a zero-inflated Poisson model analysis, you must select an input data source. You also must assign a variable to the Dependent variable role.

To filter the input data source, click .

Zero-inflated Poisson Model: Setting the Model Options

To create a zero-inflated Poisson model:

1. From the Model type drop-down list, select Zero-inflated Poisson.
2. Specify the effects for the model.

   You can display the main effects model or create a custom model. To create a custom model, select the Custom Model option, and then click Edit. The Model Effects Builder opens. All continuous variables and categorical variables are listed in the left pane.

   When you finish, click OK. The effects that you specified now appear on the Model tab.

3. Specify the link function to use to compute the probability of zeros. You can choose from the logistic function (default) or the normal function.

Setting the Options

Zero-inflated Poisson Model: Creating Output Data

You can create the following output:

• a table that contains the default statistics from the analysis and any selected statistics, such as the probability of the dependent variable taking the current value and the linear predictor. When naming this table, you must specify a CAS engine libref.
• a data set of parameter estimates. When naming this data set, you must specify a SAS libref.