OHEEncoder

class lightautoml.transformers.categorical.OHEEncoder(make_sparse=None, total_feats_cnt=None, dtype=numpy.float32)

Bases: LAMLTransformer

Simple OneHotEncoder over label encoded categories.

Parameters:

• make_sparse (Optional[bool]) – Create sparse matrix.

• total_feats_cnt (Optional[int]) – Initial features number.

• dtype (type) – Dtype of new features.

property features

Features list.

fit(dataset)

Calc output shapes.

Automatically do ohe in sparse form if approximate fill_rate < 0.2.

Parameters:

dataset (Union[NumpyDataset, PandasDataset]) – Pandas or Numpy dataset of categorical features.

Returns:

self.

transform(dataset)

Transform categorical dataset to ohe.

Parameters:

dataset (Union[NumpyDataset, PandasDataset]) – Pandas or Numpy dataset of categorical features.

Return type:

Union[NumpyDataset, CSRSparseDataset]

Returns:

Numpy dataset with encoded labels.