scikit-learn/sklearn/feature_selection/from_model.py

259 lines
9.3 KiB
Python

# Authors: Gilles Louppe, Mathieu Blondel, Maheshakya Wijewardena
# License: BSD 3 clause
import numpy as np
from .base import SelectorMixin
from ..base import TransformerMixin, BaseEstimator, clone
from ..externals import six
from ..utils import safe_mask, check_array, deprecated
from ..utils.validation import check_is_fitted
from ..exceptions import NotFittedError
def _get_feature_importances(estimator):
"""Retrieve or aggregate feature importances from estimator"""
if hasattr(estimator, "feature_importances_"):
importances = estimator.feature_importances_
elif hasattr(estimator, "coef_"):
if estimator.coef_.ndim == 1:
importances = np.abs(estimator.coef_)
else:
importances = np.sum(np.abs(estimator.coef_), axis=0)
else:
raise ValueError(
"The underlying estimator %s has no `coef_` or "
"`feature_importances_` attribute. Either pass a fitted estimator"
" to SelectFromModel or call fit before calling transform."
% estimator.__class__.__name__)
return importances
def _calculate_threshold(estimator, importances, threshold):
"""Interpret the threshold value"""
if threshold is None:
# determine default from estimator
est_name = estimator.__class__.__name__
if ((hasattr(estimator, "penalty") and estimator.penalty == "l1") or
"Lasso" in est_name):
# the natural default threshold is 0 when l1 penalty was used
threshold = 1e-5
else:
threshold = "mean"
if isinstance(threshold, six.string_types):
if "*" in threshold:
scale, reference = threshold.split("*")
scale = float(scale.strip())
reference = reference.strip()
if reference == "median":
reference = np.median(importances)
elif reference == "mean":
reference = np.mean(importances)
else:
raise ValueError("Unknown reference: " + reference)
threshold = scale * reference
elif threshold == "median":
threshold = np.median(importances)
elif threshold == "mean":
threshold = np.mean(importances)
else:
raise ValueError("Expected threshold='mean' or threshold='median' "
"got %s" % threshold)
else:
threshold = float(threshold)
return threshold
class _LearntSelectorMixin(TransformerMixin):
# Note because of the extra threshold parameter in transform, this does
# not naturally extend from SelectorMixin
"""Transformer mixin selecting features based on importance weights.
This implementation can be mixin on any estimator that exposes a
``feature_importances_`` or ``coef_`` attribute to evaluate the relative
importance of individual features for feature selection.
"""
@deprecated('Support to use estimators as feature selectors will be '
'removed in version 0.19. Use SelectFromModel instead.')
def transform(self, X, threshold=None):
"""Reduce X to its most important features.
Uses ``coef_`` or ``feature_importances_`` to determine the most
important features. For models with a ``coef_`` for each class, the
absolute sum over the classes is used.
Parameters
----------
X : array or scipy sparse matrix of shape [n_samples, n_features]
The input samples.
threshold : string, float or None, optional (default=None)
The threshold value to use for feature selection. Features whose
importance is greater or equal are kept while the others are
discarded. If "median" (resp. "mean"), then the threshold value is
the median (resp. the mean) of the feature importances. A scaling
factor (e.g., "1.25*mean") may also be used. If None and if
available, the object attribute ``threshold`` is used. Otherwise,
"mean" is used by default.
Returns
-------
X_r : array of shape [n_samples, n_selected_features]
The input samples with only the selected features.
"""
check_is_fitted(self, ('coef_', 'feature_importances_'),
all_or_any=any)
X = check_array(X, 'csc')
importances = _get_feature_importances(self)
if len(importances) != X.shape[1]:
raise ValueError("X has different number of features than"
" during model fitting.")
if threshold is None:
threshold = getattr(self, 'threshold', None)
threshold = _calculate_threshold(self, importances, threshold)
# Selection
try:
mask = importances >= threshold
except TypeError:
# Fails in Python 3.x when threshold is str;
# result is array of True
raise ValueError("Invalid threshold: all features are discarded.")
if np.any(mask):
mask = safe_mask(X, mask)
return X[:, mask]
else:
raise ValueError("Invalid threshold: all features are discarded.")
class SelectFromModel(BaseEstimator, SelectorMixin):
"""Meta-transformer for selecting features based on importance weights.
.. versionadded:: 0.17
Parameters
----------
estimator : object
The base estimator from which the transformer is built.
This can be both a fitted (if ``prefit`` is set to True)
or a non-fitted estimator.
threshold : string, float, optional default None
The threshold value to use for feature selection. Features whose
importance is greater or equal are kept while the others are
discarded. If "median" (resp. "mean"), then the ``threshold`` value is
the median (resp. the mean) of the feature importances. A scaling
factor (e.g., "1.25*mean") may also be used. If None and if the
estimator has a parameter penalty set to l1, either explicitly
or implicitly (e.g, Lasso), the threshold used is 1e-5.
Otherwise, "mean" is used by default.
prefit : bool, default False
Whether a prefit model is expected to be passed into the constructor
directly or not. If True, ``transform`` must be called directly
and SelectFromModel cannot be used with ``cross_val_score``,
``GridSearchCV`` and similar utilities that clone the estimator.
Otherwise train the model using ``fit`` and then ``transform`` to do
feature selection.
Attributes
----------
`estimator_`: an estimator
The base estimator from which the transformer is built.
This is stored only when a non-fitted estimator is passed to the
``SelectFromModel``, i.e when prefit is False.
`threshold_`: float
The threshold value used for feature selection.
"""
def __init__(self, estimator, threshold=None, prefit=False):
self.estimator = estimator
self.threshold = threshold
self.prefit = prefit
def _get_support_mask(self):
# SelectFromModel can directly call on transform.
if self.prefit:
estimator = self.estimator
elif hasattr(self, 'estimator_'):
estimator = self.estimator_
else:
raise ValueError(
'Either fit the model before transform or set "prefit=True"'
' while passing the fitted estimator to the constructor.')
scores = _get_feature_importances(estimator)
self.threshold_ = _calculate_threshold(estimator, scores,
self.threshold)
return scores >= self.threshold_
def fit(self, X, y=None, **fit_params):
"""Fit the SelectFromModel meta-transformer.
Parameters
----------
X : array-like of shape (n_samples, n_features)
The training input samples.
y : array-like, shape (n_samples,)
The target values (integers that correspond to classes in
classification, real numbers in regression).
**fit_params : Other estimator specific parameters
Returns
-------
self : object
Returns self.
"""
if self.prefit:
raise NotFittedError(
"Since 'prefit=True', call transform directly")
if not hasattr(self, "estimator_"):
self.estimator_ = clone(self.estimator)
self.estimator_.fit(X, y, **fit_params)
return self
def partial_fit(self, X, y=None, **fit_params):
"""Fit the SelectFromModel meta-transformer only once.
Parameters
----------
X : array-like of shape (n_samples, n_features)
The training input samples.
y : array-like, shape (n_samples,)
The target values (integers that correspond to classes in
classification, real numbers in regression).
**fit_params : Other estimator specific parameters
Returns
-------
self : object
Returns self.
"""
if self.prefit:
raise NotFittedError(
"Since 'prefit=True', call transform directly")
if not hasattr(self, "estimator_"):
self.estimator_ = clone(self.estimator)
self.estimator_.partial_fit(X, y, **fit_params)
return self