col_gen_estimator.Split

class col_gen_estimator.Split

All split related information.

Attributes:

id: (int) ID of the split.

feature: (int) Index of the feature used for the split.

threshold: (float) Threshold value for the split. The row takes left branch

if the feature <= threshold.

left_rows: (set(int)) IDs of the rows that take left branch on this split.

This is used for faster computing of the rows that follow a particular path.

right_rows: (set(int)) IDs of the rows that take right branch on this

split.

left_cuts: (set(int)) Indices of the cuts added as (beta) constraints that

correspopnd to a row that take left branch on this split. This is used for faster computing of the cuts that follow a particular path.

right_cuts: (set(int)) Indices of the cuts added as (beta) constraints that

correspopnd to a row that take left branch on this split.

__init__() → None

Examples using col_gen_estimator.Split

sphx_glr_auto_examples_dtree_experiment.py

Code to launch decision tree experiments.