Transformer

IndexSubset

Index subsetting transformer.

In transform, subsets X to the indices in y.index. If y is None, returns X without subsetting. numpy-based X are interpreted as having a RangeIndex starting at n, where n is the number of numpy rows seen so far through fit and update. Non-pandas types are interpreted as having index as after conversion to pandas, via datatypes.convert_to, to the "pd.DataFrame" sktime type.

Quickstart

python

from sktime.transformations.subset import IndexSubset

estimator = IndexSubset(index_treatment='keep')

Parameters(1)

index_treatmentstr, optional, one of “keep” (default) or “remove”: determines which indices are kept in Xt = transform(X, y) “keep” = all indices in y also appear in Xt. If not present in X, NA is filled. “remove” = only indices that appear in both X and y are present in Xt.

Examples

>>> from sktime.transformations.subset import IndexSubset
>>> from sktime.datasets import load_airline
>>> X = load_airline ()[0: 32 ]
>>> y = load_airline ()[24: 42 ]
>>> transformer = IndexSubset ()
>>> X_subset = transformer. fit_transform (X = X, y = y)