TrainTestLabelDrift.run_logic#

TrainTestLabelDrift.run_logic(context: Context) CheckResult[source]#

Calculate drift for all columns.

Returns
CheckResult

value: drift score. display: label distribution graph, comparing the train and test distributions.