Avoid aggregation if impurity is 0.0 | 59 ms | Passed |
Avoid aggregation on the last level | 61 ms | Passed |
Binary classification with 3-ary (ordered) categorical features, with no samples for one category: split calculation | 19 ms | Passed |
Binary classification with binary (ordered) categorical features: split calculation | 18 ms | Passed |
Binary classification with continuous features and node Id cache: subsampling features | 0.84 sec | Passed |
Binary classification with continuous features: split calculation | 55 ms | Passed |
Binary classification with continuous features: subsampling features | 0.87 sec | Passed |
Multiclass classification with ordered categorical features: split calculations | 27 ms | Passed |
Multiclass classification with unordered categorical features: split calculations | 19 ms | Passed |
SPARK-3159 tree model redundancy - classification | 0.2 sec | Passed |
SPARK-3159 tree model redundancy - regression | 0.2 sec | Passed |
Second level node building with vs. without groups | 0.24 sec | Passed |
Use soft prediction for binary classification with ordered categorical features | 42 ms | Passed |
computeFeatureImportance, featureImportances | 2 ms | Passed |
extract categories from a number for multiclass classification | 0 ms | Passed |
find splits for a continuous feature | 47 ms | Passed |
minWeightFraction and minInstancesPerNode | 0.26 sec | Passed |
normalizeMapValues | 0 ms | Passed |
train with constant features | 0.12 sec | Passed |
train with empty arrays | 13 ms | Passed |
weights at arbitrary scale | 0.26 sec | Passed |