skada.per_domain_split

skada.per_domain_split(*arrays, sample_domain)

Split data into multiple source and target domains

Parameters:

*arrayssequence of array-like of identical shape (n_samples, n_features)

Input features and target variable(s), and or sample_weights to be split. All arrays should have the same length except if None is given then a couple of None variables are returned to allow for optional sample_weight.

sample_domainarray-like of shape (n_samples,)

Array specifying the domain labels for each sample.

split_source_targetbool, optional (default=False)

Whether to split the source and target domains.

Returns:

source_dictdict

dict of source domains (contain tuple of subset of arrays).

target_dictdict

dict of target domains (contain tuple of subset of arrays).