tabensemb.data.utils#

Functions

fill_cat_nan(df, cat_dtypes)

Imputation of categorical features.

get_corr_sets(where_corr, names)

get_imputed_dtype(dtype)

Numerical columns will be transformed to "int", and others will be transformed to "str".

get_unknown_value(dtype)

Select the unknown value for the dtype judged by get_imputed_dtype().

Classes

OrdinalEncoder()

An ordinal encoder for categorical features that better supports pd.DataFrame even with missing columns.