ncem.estimators.EstimatorInteractions.split_data_target_cell

EstimatorInteractions.split_data_target_cell(target_cell, test_split, validation_split, seed=1)

Split nodes randomly into partitions.

Parameters

• target_cell (str) – Target cell type.

• test_split (float) – Fraction of total nodes to be in test set.

• validation_split (float) – Fraction of train-eval nodes to be in validation split.

• seed (int) – Seed for random selection of observations.

Raises

ValueError – If evaluation or test dataset are empty.