# Copyright 2022 The SQLNet Company GmbH
#
# This file is licensed under the Elastic License 2.0 (ELv2).
# Refer to the LICENSE.txt file in the root of the repository
# for details.
#
"""
Contains routines for preprocessing data frames.
"""
from dataclasses import dataclass
from .preprocessor import _Preprocessor
from .validate import _validate
[docs]@dataclass(repr=False)
class CategoryTrimmer(_Preprocessor):
"""
Reduces the cardinality of high-cardinality categorical columns.
Args:
max_num_categories (int, optional):
The maximum cardinality allowed. If the cardinality is
higher than that only the most frequent categories will
be kept, all others will be trimmed.
min_freq (int, optional):
The minimum frequency required for a category to be
included.
Example:
.. code-block:: python
category_trimmer = getml.preprocessors.CategoryTrimmer()
pipe = getml.Pipeline(
population=population_placeholder,
peripheral=[order_placeholder, trans_placeholder],
preprocessors=[category_trimmer],
feature_learners=[feature_learner_1, feature_learner_2],
feature_selectors=feature_selector,
predictors=predictor,
share_selected_features=0.5
)
"""
max_num_categories: int = 999
min_freq: int = 30
[docs] def validate(self, params=None):
"""Checks both the types and the values of all instance
variables and raises an exception if something is off.
Args:
params (dict, optional):
A dictionary containing
the parameters to validate. If not is passed,
the own parameters will be validated.
"""
params = _validate(self, params)
if not isinstance(params["max_num_categories"], int):
raise TypeError("'max_num_categories' must be an int.")
if not isinstance(params["min_freq"], int):
raise TypeError("'min_freq' must be an int.")
if params["max_num_categories"] < 0:
raise ValueError("'max_num_categories' cannot be negative.")
if params["min_freq"] < 0:
raise ValueError("'min_freq' cannot be negative.")