LabelEncoder

class lightautoml.transformers.categorical.LabelEncoder(subs=None, random_state=42)[source]

Simple LabelEncoder in order of frequency.

Labels are integers from 1 to n. Unknown category encoded as 0. NaN is handled as a category value.

Parameters:

subs (Optional[int]) – Subsample to calculate freqs. If None - full data.
random_state (int) – Random state to take subsample.

fit(dataset)[source]

Estimate label frequencies and create encoding dicts.

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

transform(dataset)[source]

Transform categorical dataset to int labels.

Parameters:: dataset (Union[NumpyDataset, PandasDataset]) – Pandas or Numpy dataset of categorical features.
Return type:: NumpyDataset
Returns:: Numpy dataset with encoded labels.