Skip to main content


         This documentation site is for previous versions. Visit our new documentation site for current releases.      
 

Managing grouping options for symbolic predictors manually

Updated on July 5, 2022

Combine values or ranges based on the similarity of symbolic predictor behavior.

Note: You can use the Auto grouping option and enable Prediction Studio to group predictors automatically. See Auto grouping option for predictors.
  1. In the Data analysis step, click a symbolic predictor.
  2. In the predictor workspace, click the Grouping tab.
  3. In the Granularity field, enter the maximum probability of spurious difference in behavior.
    Reducing this probability reduces interval or category numbers.
  4. In the Minimum size (% of samples) field, enter the minimum number of cases in each interval or category. Click Apply.
    Use it to ensure that the behavior evidence is sufficient to be used in grouping. Intervals with few cases are combined with their nearest neighbor. Categories with few cases are combined into a residual bin.
  5. Optional: To read the data in a graph or a table, click the Graphical view tab or the Tabular view tab
  6. Optional: Disable the Ignore ordering option.
    When this option is enabled, categories are combined with any other category most similar in behavior. After you disable it, the order of the symbolic categories is assumed to have some meaning and only the neighboring categories are grouped.
  7. Optional: Disable the Merge bins below minimum size option.
    When this option is enabled, the categories below the minimum size are grouped in a residual bin on the assumption of insufficient cases for the behavior to be the basis for grouping.
  8. Confirm your updates by clicking Submit.

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us