STATS OPTBINEX

Construct optimal bins for a categorical or continuous target

This command uses the TREE CHAID algorithm on a categorical or continuous target to group categories or ranges of predictor values that have statistically indistinguishable effects on the target. It can carry out the transformations and save that syntax for future use. It also provides value labels for the new variables. Predictor variables are analyzed one at a time. The results are similar to those from OPTIMAL BINNING for continuous variables and ADP, but has the advantage of exposing the transformations and providing labels.

---

Requirements

• IBM SPSS Statistics 18 or later and the corresponding IBM SPSS Statistics-Integration Plug-in for Python. Also requires the Decision Trees option, because it uses the TREE procedure.

---

Installation intructions

1. Open IBM SPSS Statistics
2. Navigate to Utilities -> Extension Bundles -> Download and Install Extension Bundles
3. Search for the name of the extension and click Ok. Your extension will be available.

---

License

• Apache 2.0

Contributors

• JKP, IBM SPSS