TREESPLIT Procedure

ODS Table Names

PROC TREESPLIT assigns a name to each table it creates. You can use these names to refer to the table when you use the Output Delivery System (ODS) to select tables and create output data tables. Table 5 lists these names.

Table 5: ODS Tables Produced by PROC TREESPLIT

Table Name	Description	Statement	Option
BestConfiguration	Hyperparameters and objective function values for the best configuration	AUTOTUNE	Default
ConstraintHistory	Values of the various constraints at each iteration of the EGR algorithm	MITIGATEBIAS	Default
CostComplexity	Information about the number of leaves and the error for cost-complexity pruning when validation data are used	PROC TREESPLIT	PRUNINGTABLE
CrossValidateMLFitStat	Per-fold and average assessment metrics of k-fold cross validation	CROSSVALIDATION	Default
CrossValidationValues	Results for cost-complexity pruning based on cross validation	PROC TREESPLIT	CVCC
EvaluationHistory	Values of the hyperparameters and the objective function for all configurations	AUTOTUNE	Default / EVALHISTORY=ALL
GroupMetricsHistory	Metrics for accuracy and fairness for each subpopulation over each iteration of the EGR algorithm	MITIGATEBIAS	Default
IterationHistory	Values of the model accuracy and bias metric over each iteration of the EGR algorithm	MITIGATEBIAS	Default
MitigationInformation	Information about the options that are used for bias mitigation	MITIGATEBIAS	Default
ModelInfo	Information about the modeling environment	PROC TREESPLIT	Default
NObs	Observation information	PROC TREESPLIT	Default
OutputCasTables	See the section OutputCasTables Table	PROC TREESPLIT / OUTPUT	OUTMODEL= / Default
PredName	Predicted name information for interval targets	PROC TREESPLIT	PRINTTARGET
PredProbName	Predicted probability name information for nominal targets	PROC TREESPLIT	PRINTTARGET
PredIntoName	Predicted name information for nominal targets	PROC TREESPLIT	PRINTTARGET
RBAImportance	Random branch assignment variable importance	PROC TREESPLIT	RBAIMP
ReducedError	Information about the number of leaves and the error for reduced-error pruning when validation data are used	PROC TREESPLIT	PRUNINGTABLE
Timing	Breakdown of how long bias mitigation spent on the various tasks of training models	MITIGATEBIAS	Default
TreePerformance	Fit statistics for the selected tree	PROC TREESPLIT	Default
TreePlotTable	The table that is used to generate the WholeTreePlot and ZoomedTreePlot	PROC TREESPLIT	TREEPLOT
TunerInfo	Setup values used by the tuner	AUTOTUNE	Default
TunerResults	Values of the hyperparameters and the objective function for the default configuration (Iteration 0) and up to 10 best configurations found	AUTOTUNE	Default
TunerSummary	Statistics about the tuning process	AUTOTUNE	Default
TunerTiming	Total time spent on different tasks while tuning	AUTOTUNE	Default
VariableImportance	Variable importance	PROC TREESPLIT	Default
WeightHistory	Weights that are assigned by the EGR algorithm for each level of the response (target) variable and for each subpopulation	MITIGATEBIAS	Default

Last updated: June 22, 2026