Defining Custom Splits

The Define Split dialog box enables you to select the predictor and specify conditions for each split.

  1. In the tree builder, select a node on the Viewer tab, and from the menus choose:

    Tree > Grow Branch with Custom Split

  2. Select the desired predictor from the drop-down list, or click on the Predictors button to view details of each predictor. See the topic Viewing Predictor Details for more information.
  3. You can accept the default conditions for each split or select Custom to specify conditions for the split as appropriate.
    • For continuous (numeric range) predictors, you can use the Edit Range Values fields to specify the range of values that fall into each new node.
    • For categorical predictors, you can use the Edit Set Values or Edit Ordinal Values fields to specify the specific values (or range of values in case of an ordinal predictor) that map to each new node.
  4. Select Grow to regrow the branch using the selected predictor.

The tree can generally be split using any predictor, regardless of stopping rules. The only exceptions are when the node is pure (meaning that 100% of cases fall into the same target class, thus there is nothing left to split) or the chosen predictor is constant (there is nothing to split against).

Missing values into. For CHAID trees only, if missing values are available for a given predictor, you have the option when defining a custom split to assign them to a specific child node. (With C&R Tree and QUEST, missing values are handled using surrogates as defined in the algorithm. See the topic Split Details and Surrogates for more information. )