From 5c4b1bb23192a137ac22ced229c50d6b69859ac6 Mon Sep 17 00:00:00 2001
From: Andreas Mueller <t3kcit@gmail.com>
Date: Fri, 9 Dec 2016 12:43:38 -0500
Subject: [PATCH] [MRG+1] Housekeeping Deprecations for v0.19 (#7927)
* remove stuff to be removed 0.19
* more changes
* remove classes from 0.19 whatsnew
* remove _LearntSelectorMixin
* remove ProjectedGradientNMF, load_lwf_*
* minor fixes
* remove more copy from logistic regression path
* remove lda, qda from __init__.__all__
* remove pg solver in nmf from tests etc
* remove class_weight="auto" from tests
* doctest change for decision_function_shape="ovr"
* remove transfrom from tree test, minor fixes to tree tests
* some fixes in the tests
* undo changes in functions which still allow 1d input...
* also allow 1d in scale
* more test fixes...
* last test fixes in forest and tree
* svm default value change doctest failures
* pep8
* remove more class_weight="auto" stuff
* minor cosmetics in docstrings deprecated / removed behavior.
* say that store_covariance has been moved to __init__ in discriminant_analysis
---
doc/modules/classes.rst | 17 -
doc/modules/model_persistence.rst | 2 +-
doc/modules/pipeline.rst | 4 +-
doc/modules/svm.rst | 4 +-
doc/tutorial/basic/tutorial.rst | 12 +-
.../supervised_learning.rst | 2 +-
sklearn/__init__.py | 4 +-
sklearn/base.py | 8 -
.../tests/test_robust_covariance.py | 4 +-
sklearn/datasets/__init__.py | 4 -
sklearn/datasets/lfw.py | 36 --
sklearn/datasets/tests/test_lfw.py | 44 +-
sklearn/decomposition/__init__.py | 3 +-
sklearn/decomposition/nmf.py | 424 +-----------------
sklearn/decomposition/tests/test_nmf.py | 121 ++---
sklearn/discriminant_analysis.py | 51 +--
sklearn/ensemble/forest.py | 28 +-
sklearn/ensemble/gradient_boosting.py | 53 +--
sklearn/ensemble/iforest.py | 4 +-
sklearn/ensemble/tests/test_forest.py | 12 +-
.../ensemble/tests/test_gradient_boosting.py | 10 +-
sklearn/feature_selection/from_model.py | 69 +--
.../tests/test_from_model.py | 24 -
sklearn/grid_search.py | 2 +-
sklearn/lda.py | 20 -
sklearn/linear_model/base.py | 22 -
sklearn/linear_model/coordinate_descent.py | 17 +-
sklearn/linear_model/logistic.py | 40 +-
sklearn/linear_model/perceptron.py | 3 +-
sklearn/linear_model/stochastic_gradient.py | 25 +-
.../linear_model/tests/test_least_angle.py | 2 -
sklearn/linear_model/tests/test_logistic.py | 11 -
sklearn/metrics/base.py | 9 -
sklearn/metrics/regression.py | 20 +-
sklearn/model_selection/_search.py | 2 +-
sklearn/pipeline.py | 5 -
sklearn/preprocessing/data.py | 99 +---
sklearn/preprocessing/tests/test_data.py | 32 +-
sklearn/qda.py | 20 -
sklearn/svm/base.py | 26 +-
sklearn/svm/classes.py | 45 +-
sklearn/svm/tests/test_svm.py | 19 -
sklearn/tests/test_discriminant_analysis.py | 37 --
sklearn/tests/test_multiclass.py | 2 -
sklearn/tests/test_pipeline.py | 9 -
sklearn/tree/tests/test_tree.py | 48 +-
sklearn/tree/tree.py | 5 +-
sklearn/utils/__init__.py | 11 +-
sklearn/utils/class_weight.py | 26 +-
sklearn/utils/estimator_checks.py | 50 +--
sklearn/utils/testing.py | 24 -
sklearn/utils/tests/test_class_weight.py | 69 +--
sklearn/utils/tests/test_validation.py | 9 -
sklearn/utils/validation.py | 47 +-
54 files changed, 234 insertions(+), 1462 deletions(-)
delete mode 100644 sklearn/lda.py
delete mode 100644 sklearn/qda.py
diff --git a/doc/modules/classes.rst b/doc/modules/classes.rst
index 5b44889bfa..78c2e1333d 100644
--- a/doc/modules/classes.rst
+++ b/doc/modules/classes.rst
@@ -1361,23 +1361,6 @@ Low-level methods
Recently deprecated
===================
-To be removed in 0.19
----------------------
-
-.. autosummary::
- :toctree: generated/
- :template: deprecated_class.rst
-
- lda.LDA
- qda.QDA
-
-.. autosummary::
- :toctree: generated/
- :template: deprecated_function.rst
-
- datasets.load_lfw_pairs
- datasets.load_lfw_people
-
To be removed in 0.20
---------------------
diff --git a/doc/modules/model_persistence.rst b/doc/modules/model_persistence.rst
index b1903c5202..5b83bc28a7 100644
--- a/doc/modules/model_persistence.rst
+++ b/doc/modules/model_persistence.rst
@@ -23,7 +23,7 @@ persistence model, namely `pickle <https://docs.python.org/2/library/pickle.html
>>> X, y = iris.data, iris.target
>>> clf.fit(X, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
diff --git a/doc/modules/pipeline.rst b/doc/modules/pipeline.rst
index 27fd0ce94f..5ce5386343 100644
--- a/doc/modules/pipeline.rst
+++ b/doc/modules/pipeline.rst
@@ -43,7 +43,7 @@ is an estimator object::
Pipeline(steps=[('reduce_dim', PCA(copy=True, iterated_power='auto',
n_components=None, random_state=None, svd_solver='auto', tol=0.0,
whiten=False)), ('clf', SVC(C=1.0, cache_size=200, class_weight=None,
- coef0=0.0, decision_function_shape=None, degree=3, gamma='auto',
+ coef0=0.0, decision_function_shape='ovr', degree=3, gamma='auto',
kernel='rbf', max_iter=-1, probability=False, random_state=None,
shrinking=True, tol=0.001, verbose=False))])
@@ -80,7 +80,7 @@ Parameters of the estimators in the pipeline can be accessed using the
Pipeline(steps=[('reduce_dim', PCA(copy=True, iterated_power='auto',
n_components=None, random_state=None, svd_solver='auto', tol=0.0,
whiten=False)), ('clf', SVC(C=10, cache_size=200, class_weight=None,
- coef0=0.0, decision_function_shape=None, degree=3, gamma='auto',
+ coef0=0.0, decision_function_shape='ovr', degree=3, gamma='auto',
kernel='rbf', max_iter=-1, probability=False, random_state=None,
shrinking=True, tol=0.001, verbose=False))])
diff --git a/doc/modules/svm.rst b/doc/modules/svm.rst
index 1a7ec4f7f5..8fb0d481eb 100644
--- a/doc/modules/svm.rst
+++ b/doc/modules/svm.rst
@@ -77,7 +77,7 @@ n_features]`` holding the training samples, and an array y of class labels
>>> clf = svm.SVC()
>>> clf.fit(X, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
@@ -516,7 +516,7 @@ test vectors must be provided.
>>> gram = np.dot(X, X.T)
>>> clf.fit(gram, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto',
+ decision_function_shape='ovr', degree=3, gamma='auto',
kernel='precomputed', max_iter=-1, probability=False,
random_state=None, shrinking=True, tol=0.001, verbose=False)
>>> # predict on training examples
diff --git a/doc/tutorial/basic/tutorial.rst b/doc/tutorial/basic/tutorial.rst
index a55311e015..89600953a8 100644
--- a/doc/tutorial/basic/tutorial.rst
+++ b/doc/tutorial/basic/tutorial.rst
@@ -180,7 +180,7 @@ the last entry of ``digits.data``::
>>> clf.fit(digits.data[:-1], digits.target[:-1]) # doctest: +NORMALIZE_WHITESPACE
SVC(C=100.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma=0.001, kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma=0.001, kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
@@ -219,7 +219,7 @@ persistence model, namely `pickle <https://docs.python.org/2/library/pickle.html
>>> X, y = iris.data, iris.target
>>> clf.fit(X, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
@@ -293,7 +293,7 @@ maintained::
>>> clf = SVC()
>>> clf.fit(iris.data, iris.target) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
@@ -302,7 +302,7 @@ maintained::
>>> clf.fit(iris.data, iris.target_names[iris.target]) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
@@ -331,7 +331,7 @@ more than once will overwrite what was learned by any previous ``fit()``::
>>> clf = SVC()
>>> clf.set_params(kernel='linear').fit(X, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='linear',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
>>> clf.predict(X_test)
@@ -339,7 +339,7 @@ more than once will overwrite what was learned by any previous ``fit()``::
>>> clf.set_params(kernel='rbf').fit(X, y) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
>>> clf.predict(X_test)
diff --git a/doc/tutorial/statistical_inference/supervised_learning.rst b/doc/tutorial/statistical_inference/supervised_learning.rst
index eaf4dc49a2..0440a80340 100644
--- a/doc/tutorial/statistical_inference/supervised_learning.rst
+++ b/doc/tutorial/statistical_inference/supervised_learning.rst
@@ -455,7 +455,7 @@ classification --:class:`SVC` (Support Vector Classification).
>>> svc = svm.SVC(kernel='linear')
>>> svc.fit(iris_X_train, iris_y_train) # doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='linear',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
diff --git a/sklearn/__init__.py b/sklearn/__init__.py
index 4e80774bc3..26c4fff59d 100644
--- a/sklearn/__init__.py
+++ b/sklearn/__init__.py
@@ -62,10 +62,10 @@ else:
'ensemble', 'exceptions', 'externals', 'feature_extraction',
'feature_selection', 'gaussian_process', 'grid_search',
'isotonic', 'kernel_approximation', 'kernel_ridge',
- 'lda', 'learning_curve', 'linear_model', 'manifold', 'metrics',
+ 'learning_curve', 'linear_model', 'manifold', 'metrics',
'mixture', 'model_selection', 'multiclass', 'multioutput',
'naive_bayes', 'neighbors', 'neural_network', 'pipeline',
- 'preprocessing', 'qda', 'random_projection', 'semi_supervised',
+ 'preprocessing', 'random_projection', 'semi_supervised',
'svm', 'tree', 'discriminant_analysis',
# Non-modules:
'clone']
diff --git a/sklearn/base.py b/sklearn/base.py
index 5d26d7f8e5..1b79841746 100644
--- a/sklearn/base.py
+++ b/sklearn/base.py
@@ -10,17 +10,9 @@ import numpy as np
from scipy import sparse
from .externals import six
from .utils.fixes import signature
-from .utils.deprecation import deprecated
-from .exceptions import ChangedBehaviorWarning as _ChangedBehaviorWarning
from . import __version__
-@deprecated("ChangedBehaviorWarning has been moved into the sklearn.exceptions"
- " module. It will not be available here from version 0.19")
-class ChangedBehaviorWarning(_ChangedBehaviorWarning):
- pass
-
-
##############################################################################
def _first_and_last_element(arr):
"""Returns first and last element of numpy array or sparse matrix."""
diff --git a/sklearn/covariance/tests/test_robust_covariance.py b/sklearn/covariance/tests/test_robust_covariance.py
index be5b65cd76..27e423b410 100644
--- a/sklearn/covariance/tests/test_robust_covariance.py
+++ b/sklearn/covariance/tests/test_robust_covariance.py
@@ -44,14 +44,14 @@ def test_mcd():
def test_fast_mcd_on_invalid_input():
X = np.arange(100)
- assert_raise_message(ValueError, 'fast_mcd expects at least 2 samples',
+ assert_raise_message(ValueError, 'Got X with X.ndim=1',
fast_mcd, X)
def test_mcd_class_on_invalid_input():
X = np.arange(100)
mcd = MinCovDet()
- assert_raise_message(ValueError, 'MinCovDet expects at least 2 samples',
+ assert_raise_message(ValueError, 'Got X with X.ndim=1',
mcd.fit, X)
diff --git a/sklearn/datasets/__init__.py b/sklearn/datasets/__init__.py
index 0a8cfc62df..c38e99acd3 100644
--- a/sklearn/datasets/__init__.py
+++ b/sklearn/datasets/__init__.py
@@ -18,8 +18,6 @@ from .base import load_sample_image
from .covtype import fetch_covtype
from .kddcup99 import fetch_kddcup99
from .mlcomp import load_mlcomp
-from .lfw import load_lfw_pairs
-from .lfw import load_lfw_people
from .lfw import fetch_lfw_pairs
from .lfw import fetch_lfw_people
from .twenty_newsgroups import fetch_20newsgroups
@@ -74,8 +72,6 @@ __all__ = ['clear_data_home',
'load_files',
'load_iris',
'load_breast_cancer',
- 'load_lfw_pairs',
- 'load_lfw_people',
'load_linnerud',
'load_mlcomp',
'load_sample_image',
diff --git a/sklearn/datasets/lfw.py b/sklearn/datasets/lfw.py
index e191efb37b..d155cfe478 100644
--- a/sklearn/datasets/lfw.py
+++ b/sklearn/datasets/lfw.py
@@ -26,8 +26,6 @@ detector from various online websites.
from os import listdir, makedirs, remove, rename
from os.path import join, exists, isdir
-from sklearn.utils import deprecated
-
import logging
import numpy as np
@@ -376,23 +374,6 @@ def _fetch_lfw_pairs(index_file_path, data_folder_path, slice_=None,
return pairs, target, np.array(['Different persons', 'Same person'])
-@deprecated("Function 'load_lfw_people' has been deprecated in 0.17 and will "
- "be removed in 0.19."
- "Use fetch_lfw_people(download_if_missing=False) instead.")
-def load_lfw_people(download_if_missing=False, **kwargs):
- """
- Alias for fetch_lfw_people(download_if_missing=False)
-
- .. deprecated:: 0.17
- This function will be removed in 0.19.
- Use :func:`sklearn.datasets.fetch_lfw_people` with parameter
- ``download_if_missing=False`` instead.
-
- Check fetch_lfw_people.__doc__ for the documentation and parameter list.
- """
- return fetch_lfw_people(download_if_missing=download_if_missing, **kwargs)
-
-
def fetch_lfw_pairs(subset='train', data_home=None, funneled=True, resize=0.5,
color=False, slice_=(slice(70, 195), slice(78, 172)),
download_if_missing=True):
@@ -509,20 +490,3 @@ def fetch_lfw_pairs(subset='train', data_home=None, funneled=True, resize=0.5,
return Bunch(data=pairs.reshape(len(pairs), -1), pairs=pairs,
target=target, target_names=target_names,
DESCR="'%s' segment of the LFW pairs dataset" % subset)
-
-
-@deprecated("Function 'load_lfw_pairs' has been deprecated in 0.17 and will "
- "be removed in 0.19."
- "Use fetch_lfw_pairs(download_if_missing=False) instead.")
-def load_lfw_pairs(download_if_missing=False, **kwargs):
- """
- Alias for fetch_lfw_pairs(download_if_missing=False)
-
- .. deprecated:: 0.17
- This function will be removed in 0.19.
- Use :func:`sklearn.datasets.fetch_lfw_pairs` with parameter
- ``download_if_missing=False`` instead.
-
- Check fetch_lfw_pairs.__doc__ for the documentation and parameter list.
- """
- return fetch_lfw_pairs(download_if_missing=download_if_missing, **kwargs)
diff --git a/sklearn/datasets/tests/test_lfw.py b/sklearn/datasets/tests/test_lfw.py
index 83f7ba16c8..3e5875a060 100644
--- a/sklearn/datasets/tests/test_lfw.py
+++ b/sklearn/datasets/tests/test_lfw.py
@@ -22,14 +22,11 @@ try:
except ImportError:
imsave = None
-from sklearn.datasets import load_lfw_pairs
-from sklearn.datasets import load_lfw_people
from sklearn.datasets import fetch_lfw_pairs
from sklearn.datasets import fetch_lfw_people
from sklearn.utils.testing import assert_array_equal
from sklearn.utils.testing import assert_equal
-from sklearn.utils.testing import assert_warns_message
from sklearn.utils.testing import SkipTest
from sklearn.utils.testing import raises
@@ -115,20 +112,14 @@ def teardown_module():
@raises(IOError)
def test_load_empty_lfw_people():
- fetch_lfw_people(data_home=SCIKIT_LEARN_EMPTY_DATA, download_if_missing=False)
-
-
-def test_load_lfw_people_deprecation():
- msg = ("Function 'load_lfw_people' has been deprecated in 0.17 and will be "
- "removed in 0.19."
- "Use fetch_lfw_people(download_if_missing=False) instead.")
- assert_warns_message(DeprecationWarning, msg, load_lfw_people,
- data_home=SCIKIT_LEARN_DATA)
+ fetch_lfw_people(data_home=SCIKIT_LEARN_EMPTY_DATA,
+ download_if_missing=False)
def test_load_fake_lfw_people():
lfw_people = fetch_lfw_people(data_home=SCIKIT_LEARN_DATA,
- min_faces_per_person=3, download_if_missing=False)
+ min_faces_per_person=3,
+ download_if_missing=False)
# The data is croped around the center as a rectangular bounding box
# around the face. Colors are converted to gray levels:
@@ -144,8 +135,9 @@ def test_load_fake_lfw_people():
# It is possible to ask for the original data without any croping or color
# conversion and not limit on the number of picture per person
- lfw_people = fetch_lfw_people(data_home=SCIKIT_LEARN_DATA,
- resize=None, slice_=None, color=True, download_if_missing=False)
+ lfw_people = fetch_lfw_people(data_home=SCIKIT_LEARN_DATA, resize=None,
+ slice_=None, color=True,
+ download_if_missing=False)
assert_equal(lfw_people.images.shape, (17, 250, 250, 3))
# the ids and class names are the same as previously
@@ -158,24 +150,19 @@ def test_load_fake_lfw_people():
@raises(ValueError)
def test_load_fake_lfw_people_too_restrictive():
- fetch_lfw_people(data_home=SCIKIT_LEARN_DATA, min_faces_per_person=100, download_if_missing=False)
+ fetch_lfw_people(data_home=SCIKIT_LEARN_DATA, min_faces_per_person=100,
+ download_if_missing=False)
@raises(IOError)
def test_load_empty_lfw_pairs():
- fetch_lfw_pairs(data_home=SCIKIT_LEARN_EMPTY_DATA, download_if_missing=False)
-
-
-def test_load_lfw_pairs_deprecation():
- msg = ("Function 'load_lfw_pairs' has been deprecated in 0.17 and will be "
- "removed in 0.19."
- "Use fetch_lfw_pairs(download_if_missing=False) instead.")
- assert_warns_message(DeprecationWarning, msg, load_lfw_pairs,
- data_home=SCIKIT_LEARN_DATA)
+ fetch_lfw_pairs(data_home=SCIKIT_LEARN_EMPTY_DATA,
+ download_if_missing=False)
def test_load_fake_lfw_pairs():
- lfw_pairs_train = fetch_lfw_pairs(data_home=SCIKIT_LEARN_DATA, download_if_missing=False)
+ lfw_pairs_train = fetch_lfw_pairs(data_home=SCIKIT_LEARN_DATA,
+ download_if_missing=False)
# The data is croped around the center as a rectangular bounding box
# around the face. Colors are converted to gray levels:
@@ -190,8 +177,9 @@ def test_load_fake_lfw_pairs():
# It is possible to ask for the original data without any croping or color
# conversion
- lfw_pairs_train = fetch_lfw_pairs(data_home=SCIKIT_LEARN_DATA,
- resize=None, slice_=None, color=True, download_if_missing=False)
+ lfw_pairs_train = fetch_lfw_pairs(data_home=SCIKIT_LEARN_DATA, resize=None,
+ slice_=None, color=True,
+ download_if_missing=False)
assert_equal(lfw_pairs_train.pairs.shape, (10, 2, 250, 250, 3))
# the ids and class names are the same as previously
diff --git a/sklearn/decomposition/__init__.py b/sklearn/decomposition/__init__.py
index 24690c8c8f..faca56b91b 100644
--- a/sklearn/decomposition/__init__.py
+++ b/sklearn/decomposition/__init__.py
@@ -4,7 +4,7 @@ algorithms, including among others PCA, NMF or ICA. Most of the algorithms of
this module can be regarded as dimensionality reduction techniques.
"""
-from .nmf import NMF, ProjectedGradientNMF, non_negative_factorization
+from .nmf import NMF, non_negative_factorization
from .pca import PCA, RandomizedPCA
from .incremental_pca import IncrementalPCA
from .kernel_pca import KernelPCA
@@ -26,7 +26,6 @@ __all__ = ['DictionaryLearning',
'MiniBatchSparsePCA',
'NMF',
'PCA',
- 'ProjectedGradientNMF',
'RandomizedPCA',
'SparseCoder',
'SparsePCA',
diff --git a/sklearn/decomposition/nmf.py b/sklearn/decomposition/nmf.py
index cf5fc431e6..3b71079d99 100644
--- a/sklearn/decomposition/nmf.py
+++ b/sklearn/decomposition/nmf.py
@@ -24,7 +24,6 @@ from ..utils import check_random_state, check_array
from ..utils.extmath import randomized_svd, safe_sparse_dot, squared_norm
from ..utils.extmath import fast_dot
from ..utils.validation import check_is_fitted, check_non_negative
-from ..utils import deprecated
from ..exceptions import ConvergenceWarning
from .cdnmf_fast import _update_cdnmf_fast
@@ -52,12 +51,6 @@ def trace_dot(X, Y):
return np.dot(X.ravel(), Y.ravel())
-def _sparseness(x):
- """Hoyer's measure of sparsity for a vector"""
- sqrt_n = np.sqrt(len(x))
- return (sqrt_n - np.linalg.norm(x, 1) / norm(x)) / (sqrt_n - 1)
-
-
def _check_init(A, shape, whom):
A = check_array(A)
if np.shape(A) != shape:
@@ -80,20 +73,6 @@ def _safe_compute_error(X, W, H):
return error
-def _check_string_param(sparseness, solver):
- allowed_sparseness = (None, 'data', 'components')
- if sparseness not in allowed_sparseness:
- raise ValueError(
- 'Invalid sparseness parameter: got %r instead of one of %r' %
- (sparseness, allowed_sparseness))
-
- allowed_solver = ('pg', 'cd')
- if solver not in allowed_solver:
- raise ValueError(
- 'Invalid solver parameter: got %r instead of one of %r' %
- (solver, allowed_solver))
-
-
def _initialize_nmf(X, n_components, init=None, eps=1e-6,
random_state=None):
"""Algorithms for NMF initialization.
@@ -345,115 +324,6 @@ def _nls_subproblem(V, W, H, tol, max_iter, alpha=0., l1_ratio=0.,
return H, grad, n_iter
-def _update_projected_gradient_w(X, W, H, tolW, nls_max_iter, alpha, l1_ratio,
- sparseness, beta, eta):
- """Helper function for _fit_projected_gradient"""
- n_samples, n_features = X.shape
- n_components_ = H.shape[0]
-
- if sparseness is None:
- Wt, gradW, iterW = _nls_subproblem(X.T, H.T, W.T, tolW, nls_max_iter,
- alpha=alpha, l1_ratio=l1_ratio)
- elif sparseness == 'data':
- Wt, gradW, iterW = _nls_subproblem(
- safe_vstack([X.T, np.zeros((1, n_samples))]),
- safe_vstack([H.T, np.sqrt(beta) * np.ones((1,
- n_components_))]),
- W.T, tolW, nls_max_iter, alpha=alpha, l1_ratio=l1_ratio)
- elif sparseness == 'components':
- Wt, gradW, iterW = _nls_subproblem(
- safe_vstack([X.T,
- np.zeros((n_components_, n_samples))]),
- safe_vstack([H.T,
- np.sqrt(eta) * np.eye(n_components_)]),
- W.T, tolW, nls_max_iter, alpha=alpha, l1_ratio=l1_ratio)
-
- return Wt.T, gradW.T, iterW
-
-
-def _update_projected_gradient_h(X, W, H, tolH, nls_max_iter, alpha, l1_ratio,
- sparseness, beta, eta):
- """Helper function for _fit_projected_gradient"""
- n_samples, n_features = X.shape
- n_components_ = W.shape[1]
-
- if sparseness is None:
- H, gradH, iterH = _nls_subproblem(X, W, H, tolH, nls_max_iter,
- alpha=alpha, l1_ratio=l1_ratio)
- elif sparseness == 'data':
- H, gradH, iterH = _nls_subproblem(
- safe_vstack([X, np.zeros((n_components_, n_features))]),
- safe_vstack([W,
- np.sqrt(eta) * np.eye(n_components_)]),
- H, tolH, nls_max_iter, alpha=alpha, l1_ratio=l1_ratio)
- elif sparseness == 'components':
- H, gradH, iterH = _nls_subproblem(
- safe_vstack([X, np.zeros((1, n_features))]),
- safe_vstack([W, np.sqrt(beta) * np.ones((1, n_components_))]),
- H, tolH, nls_max_iter, alpha=alpha, l1_ratio=l1_ratio)
-
- return H, gradH, iterH
-
-
-def _fit_projected_gradient(X, W, H, tol, max_iter,
- nls_max_iter, alpha, l1_ratio,
- sparseness, beta, eta):
- """Compute Non-negative Matrix Factorization (NMF) with Projected Gradient
-
- References
- ----------
- C.-J. Lin. Projected gradient methods for non-negative matrix
- factorization. Neural Computation, 19(2007), 2756-2779.
- http://www.csie.ntu.edu.tw/~cjlin/nmf/
-
- P. Hoyer. Non-negative Matrix Factorization with Sparseness Constraints.
- Journal of Machine Learning Research 2004.
- """
- gradW = (np.dot(W, np.dot(H, H.T)) -
- safe_sparse_dot(X, H.T, dense_output=True))
- gradH = (np.dot(np.dot(W.T, W), H) -
- safe_sparse_dot(W.T, X, dense_output=True))
-
- init_grad = squared_norm(gradW) + squared_norm(gradH.T)
- # max(0.001, tol) to force alternating minimizations of W and H
- tolW = max(0.001, tol) * np.sqrt(init_grad)
- tolH = tolW
-
- for n_iter in range(1, max_iter + 1):
- # stopping condition
- # as discussed in paper
- proj_grad_W = squared_norm(gradW * np.logical_or(gradW < 0, W > 0))
- proj_grad_H = squared_norm(gradH * np.logical_or(gradH < 0, H > 0))
-
- if (proj_grad_W + proj_grad_H) / init_grad < tol ** 2:
- break
-
- # update W
- W, gradW, iterW = _update_projected_gradient_w(X, W, H, tolW,
- nls_max_iter,
- alpha, l1_ratio,
- sparseness, beta, eta)
- if iterW == 1:
- tolW = 0.1 * tolW
-
- # update H
- H, gradH, iterH = _update_projected_gradient_h(X, W, H, tolH,
- nls_max_iter,
- alpha, l1_ratio,
- sparseness, beta, eta)
- if iterH == 1:
- tolH = 0.1 * tolH
-
- H[H == 0] = 0 # fix up negative zeros
-
- if n_iter == max_iter:
- W, _, _ = _update_projected_gradient_w(X, W, H, tol, nls_max_iter,
- alpha, l1_ratio, sparseness,
- beta, eta)
-
- return W, H, n_iter
-
-
def _update_coordinate_descent(X, W, Ht, l1_reg, l2_reg, shuffle,
random_state):
"""Helper function for _fit_coordinate_descent
@@ -604,8 +474,7 @@ def non_negative_factorization(X, W=None, H=None, n_components=None,
init='random', update_H=True, solver='cd',
tol=1e-4, max_iter=200, alpha=0., l1_ratio=0.,
regularization=None, random_state=None,
- verbose=0, shuffle=False, nls_max_iter=2000,
- sparseness=None, beta=1, eta=0.1):
+ verbose=0, shuffle=False):
"""Compute Non-negative Matrix Factorization (NMF)
Find two non-negative matrices (W, H) whose product approximates the non-
@@ -668,9 +537,8 @@ def non_negative_factorization(X, W=None, H=None, n_components=None,
Set to True, both W and H will be estimated from initial guesses.
Set to False, only W will be estimated.
- solver : 'pg' | 'cd'
+ solver : 'cd'
Numerical solver to use:
- 'pg' is a (deprecated) Projected Gradient solver.
'cd' is a Coordinate Descent solver.
tol : float, default: 1e-4
@@ -702,21 +570,6 @@ def non_negative_factorization(X, W=None, H=None, n_components=None,
shuffle : boolean, default: False
If true, randomize the order of coordinates in the CD solver.
- nls_max_iter : integer, default: 2000
- Number of iterations in NLS subproblem.
- Used only in the deprecated 'pg' solver.
-
- sparseness : 'data' | 'components' | None, default: None
- Where to enforce sparsity in the model.
- Used only in the deprecated 'pg' solver.
-
- beta : double, default: 1
- Degree of sparseness, if sparseness is not None. Larger values mean
- more sparseness. Used only in the deprecated 'pg' solver.
-
- eta : double, default: 0.1
- Degree of correctness to maintain, if sparsity is not None. Smaller
- values mean larger error. Used only in the deprecated 'pg' solver.
Returns
-------
@@ -743,7 +596,6 @@ def non_negative_factorization(X, W=None, H=None, n_components=None,
X = check_array(X, accept_sparse=('csr', 'csc'))
check_non_negative(X, "NMF (input X)")
- _check_string_param(sparseness, solver)
n_samples, n_features = X.shape
if n_components is None:
@@ -770,23 +622,7 @@ def non_negative_factorization(X, W=None, H=None, n_components=None,
W, H = _initialize_nmf(X, n_components, init=init,
random_state=random_state)
- if solver == 'pg':
- warnings.warn("'pg' solver will be removed in release 0.19."
- " Use 'cd' solver instead.", DeprecationWarning)
- if update_H: # fit_transform
- W, H, n_iter = _fit_projected_gradient(X, W, H, tol,
- max_iter,
- nls_max_iter,
- alpha, l1_ratio,
- sparseness,
- beta, eta)
- else: # transform
- W, H, n_iter = _update_projected_gradient_w(X, W, H,
- tol, nls_max_iter,
- alpha, l1_ratio,
- sparseness, beta,
- eta)
- elif solver == 'cd':
+ if solver == 'cd':
W, H, n_iter = _fit_coordinate_descent(X, W, H, tol,
max_iter,
alpha, l1_ratio,
@@ -856,10 +692,9 @@ class NMF(BaseEstimator, TransformerMixin):
- 'custom': use custom matrices W and H
- solver : 'pg' | 'cd'
+ solver : 'cd'
Numerical solver to use:
- 'pg' is a Projected Gradient solver (deprecated).
- 'cd' is a Coordinate Descent solver (recommended).
+ 'cd' is a Coordinate Descent solver.
.. versionadded:: 0.17
Coordinate Descent solver.
@@ -900,37 +735,6 @@ class NMF(BaseEstimator, TransformerMixin):
.. versionadded:: 0.17
*shuffle* parameter used in the Coordinate Descent solver.
- nls_max_iter : integer, default: 2000
- Number of iterations in NLS subproblem.
- Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- sparseness : 'data' | 'components' | None, default: None
- Where to enforce sparsity in the model.
- Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- beta : double, default: 1
- Degree of sparseness, if sparseness is not None. Larger values mean
- more sparseness. Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- eta : double, default: 0.1
- Degree of correctness to maintain, if sparsity is not None. Smaller
- values mean larger error. Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
Attributes
----------
@@ -952,9 +756,9 @@ class NMF(BaseEstimator, TransformerMixin):
>>> from sklearn.decomposition import NMF
>>> model = NMF(n_components=2, init='random', random_state=0)
>>> model.fit(X) #doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
- NMF(alpha=0.0, beta=1, eta=0.1, init='random', l1_ratio=0.0, max_iter=200,
- n_components=2, nls_max_iter=2000, random_state=0, shuffle=False,
- solver='cd', sparseness=None, tol=0.0001, verbose=0)
+ NMF(alpha=0.0, init='random', l1_ratio=0.0, max_iter=200,
+ n_components=2, random_state=0, shuffle=False,
+ solver='cd', tol=0.0001, verbose=0)
>>> model.components_
array([[ 2.09783018, 0.30560234],
@@ -974,10 +778,9 @@ class NMF(BaseEstimator, TransformerMixin):
computer sciences 92.3: 708-721, 2009.
"""
- def __init__(self, n_components=None, init=None, solver='cd',
- tol=1e-4, max_iter=200, random_state=None,
- alpha=0., l1_ratio=0., verbose=0, shuffle=False,
- nls_max_iter=2000, sparseness=None, beta=1, eta=0.1):
+ def __init__(self, n_components=None, init=None, solver='cd', tol=1e-4,
+ max_iter=200, random_state=None, alpha=0., l1_ratio=0.,
+ verbose=0, shuffle=False):
self.n_components = n_components
self.init = init
self.solver = solver
@@ -989,17 +792,6 @@ class NMF(BaseEstimator, TransformerMixin):
self.verbose = verbose
self.shuffle = shuffle
- if sparseness is not None:
- warnings.warn("Controlling regularization through the sparseness,"
- " beta and eta arguments is only available"
- " for 'pg' solver, which will be removed"
- " in release 0.19. Use another solver with L1 or L2"
- " regularization instead.", DeprecationWarning)
- self.nls_max_iter = nls_max_iter
- self.sparseness = sparseness
- self.beta = beta
- self.eta = eta
-
def fit_transform(self, X, y=None, W=None, H=None):
"""Learn a NMF model for the data X and returns the transformed data.
@@ -1029,13 +821,7 @@ class NMF(BaseEstimator, TransformerMixin):
tol=self.tol, max_iter=self.max_iter, alpha=self.alpha,
l1_ratio=self.l1_ratio, regularization='both',
random_state=self.random_state, verbose=self.verbose,
- shuffle=self.shuffle,
- nls_max_iter=self.nls_max_iter, sparseness=self.sparseness,
- beta=self.beta, eta=self.eta)
-
- if self.solver == 'pg':
- self.comp_sparseness_ = _sparseness(H.ravel())
- self.data_sparseness_ = _sparseness(W.ravel())
+ shuffle=self.shuffle)
self.reconstruction_err_ = _safe_compute_error(X, W, H)
@@ -1081,9 +867,7 @@ class NMF(BaseEstimator, TransformerMixin):
tol=self.tol, max_iter=self.max_iter, alpha=self.alpha,
l1_ratio=self.l1_ratio, regularization='both',
random_state=self.random_state, verbose=self.verbose,
- shuffle=self.shuffle,
- nls_max_iter=self.nls_max_iter, sparseness=self.sparseness,
- beta=self.beta, eta=self.eta)
+ shuffle=self.shuffle)
return W
@@ -1104,185 +888,3 @@ class NMF(BaseEstimator, TransformerMixin):
"""
check_is_fitted(self, 'n_components_')
return np.dot(W, self.components_)
-
-
-@deprecated("It will be removed in release 0.19. Use NMF instead."
- "'pg' solver is still available until release 0.19.")
-class ProjectedGradientNMF(NMF):
- """Non-Negative Matrix Factorization (NMF)
-
- Find two non-negative matrices (W, H) whose product approximates the non-
- negative matrix X. This factorization can be used for example for
- dimensionality reduction, source separation or topic extraction.
-
- The objective function is::
-
- 0.5 * ||X - WH||_Fro^2
- + alpha * l1_ratio * ||vec(W)||_1
- + alpha * l1_ratio * ||vec(H)||_1
- + 0.5 * alpha * (1 - l1_ratio) * ||W||_Fro^2
- + 0.5 * alpha * (1 - l1_ratio) * ||H||_Fro^2
-
- Where::
-
- ||A||_Fro^2 = \sum_{i,j} A_{ij}^2 (Frobenius norm)
- ||vec(A)||_1 = \sum_{i,j} abs(A_{ij}) (Elementwise L1 norm)
-
- The objective function is minimized with an alternating minimization of W
- and H.
-
- Read more in the :ref:`User Guide <NMF>`.
-
- Parameters
- ----------
- n_components : int or None
- Number of components, if n_components is not set all features
- are kept.
-
- init : 'random' | 'nndsvd' | 'nndsvda' | 'nndsvdar' | 'custom'
- Method used to initialize the procedure.
- Default: 'nndsvdar' if n_components < n_features, otherwise random.
- Valid options:
-
- - 'random': non-negative random matrices, scaled with:
- sqrt(X.mean() / n_components)
-
- - 'nndsvd': Nonnegative Double Singular Value Decomposition (NNDSVD)
- initialization (better for sparseness)
-
- - 'nndsvda': NNDSVD with zeros filled with the average of X
- (better when sparsity is not desired)
-
- - 'nndsvdar': NNDSVD with zeros filled with small random values
- (generally faster, less accurate alternative to NNDSVDa
- for when sparsity is not desired)
-
- - 'custom': use custom matrices W and H
-
- solver : 'pg' | 'cd'
- Numerical solver to use:
- 'pg' is a Projected Gradient solver (deprecated).
- 'cd' is a Coordinate Descent solver (recommended).
-
- .. versionadded:: 0.17
- Coordinate Descent solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver.
-
- tol : double, default: 1e-4
- Tolerance value used in stopping conditions.
-
- max_iter : integer, default: 200
- Number of iterations to compute.
-
- random_state : integer seed, RandomState instance, or None (default)
- Random number generator seed control.
-
- alpha : double, default: 0.
- Constant that multiplies the regularization terms. Set it to zero to
- have no regularization.
-
- .. versionadded:: 0.17
- *alpha* used in the Coordinate Descent solver.
-
- l1_ratio : double, default: 0.
- The regularization mixing parameter, with 0 <= l1_ratio <= 1.
- For l1_ratio = 0 the penalty is an elementwise L2 penalty
- (aka Frobenius Norm).
- For l1_ratio = 1 it is an elementwise L1 penalty.
- For 0 < l1_ratio < 1, the penalty is a combination of L1 and L2.
-
- .. versionadded:: 0.17
- Regularization parameter *l1_ratio* used in the Coordinate Descent
- solver.
-
- shuffle : boolean, default: False
- If true, randomize the order of coordinates in the CD solver.
-
- .. versionadded:: 0.17
- *shuffle* parameter used in the Coordinate Descent solver.
-
- nls_max_iter : integer, default: 2000
- Number of iterations in NLS subproblem.
- Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- sparseness : 'data' | 'components' | None, default: None
- Where to enforce sparsity in the model.
- Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- beta : double, default: 1
- Degree of sparseness, if sparseness is not None. Larger values mean
- more sparseness. Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- eta : double, default: 0.1
- Degree of correctness to maintain, if sparsity is not None. Smaller
- values mean larger error. Used only in the deprecated 'pg' solver.
-
- .. versionchanged:: 0.17
- Deprecated Projected Gradient solver. Use Coordinate Descent solver
- instead.
-
- Attributes
- ----------
- components_ : array, [n_components, n_features]
- Non-negative components of the data.
-
- reconstruction_err_ : number
- Frobenius norm of the matrix difference between
- the training data and the reconstructed data from
- the fit produced by the model. ``|| X - WH ||_2``
-
- n_iter_ : int
- Actual number of iterations.
-
- Examples
- --------
- >>> import numpy as np
- >>> X = np.array([[1,1], [2, 1], [3, 1.2], [4, 1], [5, 0.8], [6, 1]])
- >>> from sklearn.decomposition import NMF
- >>> model = NMF(n_components=2, init='random', random_state=0)
- >>> model.fit(X) #doctest: +ELLIPSIS +NORMALIZE_WHITESPACE
- NMF(alpha=0.0, beta=1, eta=0.1, init='random', l1_ratio=0.0, max_iter=200,
- n_components=2, nls_max_iter=2000, random_state=0, shuffle=False,
- solver='cd', sparseness=None, tol=0.0001, verbose=0)
-
- >>> model.components_
- array([[ 2.09783018, 0.30560234],
- [ 2.13443044, 2.13171694]])
- >>> model.reconstruction_err_ #doctest: +ELLIPSIS
- 0.00115993...
-
- References
- ----------
- C.-J. Lin. Projected gradient methods for non-negative matrix
- factorization. Neural Computation, 19(2007), 2756-2779.
- http://www.csie.ntu.edu.tw/~cjlin/nmf/
-
- Cichocki, Andrzej, and P. H. A. N. Anh-Huy. "Fast local algorithms for
- large scale nonnegative matrix and tensor factorizations."
- IEICE transactions on fundamentals of electronics, communications and
- computer sciences 92.3: 708-721, 2009.
- """
-
- def __init__(self, n_components=None, solver='pg', init=None,
- tol=1e-4, max_iter=200, random_state=None,
- alpha=0., l1_ratio=0., verbose=0,
- nls_max_iter=2000, sparseness=None, beta=1, eta=0.1):
- super(ProjectedGradientNMF, self).__init__(
- n_components=n_components, init=init, solver='pg', tol=tol,
- max_iter=max_iter, random_state=random_state, alpha=alpha,
- l1_ratio=l1_ratio, verbose=verbose, nls_max_iter=nls_max_iter,
- sparseness=sparseness, beta=beta, eta=eta)
diff --git a/sklearn/decomposition/tests/test_nmf.py b/sklearn/decomposition/tests/test_nmf.py
index c431dd3842..bb93ed94f3 100644
--- a/sklearn/decomposition/tests/test_nmf.py
+++ b/sklearn/decomposition/tests/test_nmf.py
@@ -1,7 +1,6 @@
import numpy as np
from scipy import linalg
-from sklearn.decomposition import (NMF, ProjectedGradientNMF,
- non_negative_factorization)
+from sklearn.decomposition import NMF, non_negative_factorization
from sklearn.decomposition import nmf # For testing internals
from scipy.sparse import csc_matrix
@@ -10,9 +9,7 @@ from sklearn.utils.testing import assert_false
from sklearn.utils.testing import assert_raise_message, assert_no_warnings
from sklearn.utils.testing import assert_array_almost_equal
from sklearn.utils.testing import assert_almost_equal
-from sklearn.utils.testing import assert_greater
from sklearn.utils.testing import assert_less
-from sklearn.utils.testing import ignore_warnings
from sklearn.base import clone
@@ -27,16 +24,13 @@ def test_initialize_nn_output():
assert_false((W < 0).any() or (H < 0).any())
-@ignore_warnings
def test_parameter_checking():
A = np.ones((2, 2))
name = 'spam'
- msg = "Invalid solver parameter: got 'spam' instead of one of"
+ msg = "Invalid solver parameter 'spam'"
assert_raise_message(ValueError, msg, NMF(solver=name).fit, A)
msg = "Invalid init parameter: got 'spam' instead of one of"
assert_raise_message(ValueError, msg, NMF(init=name).fit, A)
- msg = "Invalid sparseness parameter: got 'spam' instead of one of"
- assert_raise_message(ValueError, msg, NMF(sparseness=name).fit, A)
msg = "Negative values in data passed to"
assert_raise_message(ValueError, msg, NMF().fit, -A)
@@ -71,27 +65,22 @@ def test_initialize_variants():
assert_almost_equal(evl[ref != 0], ref[ref != 0])
-@ignore_warnings
def test_nmf_fit_nn_output():
# Test that the decomposition does not contain negative values
A = np.c_[5 * np.ones(5) - np.arange(1, 6),
5 * np.ones(5) + np.arange(1, 6)]
- for solver in ('pg', 'cd'):
- for init in (None, 'nndsvd', 'nndsvda', 'nndsvdar'):
- model = NMF(n_components=2, solver=solver, init=init,
- random_state=0)
- transf = model.fit_transform(A)
- assert_false((model.components_ < 0).any() or
- (transf < 0).any())
+ for init in (None, 'nndsvd', 'nndsvda', 'nndsvdar'):
+ model = NMF(n_components=2, init=init, random_state=0)
+ transf = model.fit_transform(A)
+ assert_false((model.components_ < 0).any() or
+ (transf < 0).any())
-@ignore_warnings
def test_nmf_fit_close():
# Test that the fit is not too far away
- for solver in ('pg', 'cd'):
- pnmf = NMF(5, solver=solver, init='nndsvd', random_state=0)
- X = np.abs(random_state.randn(6, 5))
- assert_less(pnmf.fit(X).reconstruction_err_, 0.05)
+ pnmf = NMF(5, init='nndsvd', random_state=0)
+ X = np.abs(random_state.randn(6, 5))
+ assert_less(pnmf.fit(X).reconstruction_err_, 0.05)
def test_nls_nn_output():
@@ -109,15 +98,13 @@ def test_nls_close():
assert_true((np.abs(Ap - A) < 0.01).all())
-@ignore_warnings
def test_nmf_transform():
# Test that NMF.transform returns close values
A = np.abs(random_state.randn(6, 5))
- for solver in ('pg', 'cd'):
- m = NMF(solver=solver, n_components=4, init='nndsvd', random_state=0)
- ft = m.fit_transform(A)
- t = m.transform(A)
- assert_array_almost_equal(ft, t, decimal=2)
+ m = NMF(n_components=4, init='nndsvd', random_state=0)
+ ft = m.fit_transform(A)
+ t = m.transform(A)
+ assert_array_almost_equal(ft, t, decimal=2)
def test_nmf_transform_custom_init():
@@ -134,45 +121,23 @@ def test_nmf_transform_custom_init():
m.transform(A)
-@ignore_warnings
def test_nmf_inverse_transform():
# Test that NMF.inverse_transform returns close values
random_state = np.random.RandomState(0)
A = np.abs(random_state.randn(6, 4))
- for solver in ('pg', 'cd'):
- m = NMF(solver=solver, n_components=4, init='random', random_state=0)
- m.fit_transform(A)
- t = m.transform(A)
- A_new = m.inverse_transform(t)
- assert_array_almost_equal(A, A_new, decimal=2)
+ m = NMF(n_components=4, init='random', random_state=0)
+ m.fit_transform(A)
+ t = m.transform(A)
+ A_new = m.inverse_transform(t)
+ assert_array_almost_equal(A, A_new, decimal=2)
-@ignore_warnings
def test_n_components_greater_n_features():
# Smoke test for the case of more components than features.
A = np.abs(random_state.randn(30, 10))
NMF(n_components=15, random_state=0, tol=1e-2).fit(A)
-@ignore_warnings
-def test_projgrad_nmf_sparseness():
- # Test sparseness
- # Test that sparsity constraints actually increase sparseness in the
- # part where they are applied.
- tol = 1e-2
- A = np.abs(random_state.randn(10, 10))
- m = ProjectedGradientNMF(n_components=5, random_state=0, tol=tol).fit(A)
- data_sp = ProjectedGradientNMF(n_components=5, sparseness='data',
- random_state=0,
- tol=tol).fit(A).data_sparseness_
- comp_sp = ProjectedGradientNMF(n_components=5, sparseness='components',
- random_state=0,
- tol=tol).fit(A).comp_sparseness_
- assert_greater(data_sp, m.data_sparseness_)
- assert_greater(comp_sp, m.comp_sparseness_)
-
-
-@ignore_warnings
def test_sparse_input():
# Test that sparse matrices are accepted as input
from scipy.sparse import csc_matrix
@@ -181,21 +146,18 @@ def test_sparse_input():
A[:, 2 * np.arange(5)] = 0
A_sparse = csc_matrix(A)
- for solver in ('pg', 'cd'):
- est1 = NMF(solver=solver, n_components=5, init='random',
- random_state=0, tol=1e-2)
- est2 = clone(est1)
+ est1 = NMF(n_components=5, init='random', random_state=0, tol=1e-2)
+ est2 = clone(est1)
- W1 = est1.fit_transform(A)
- W2 = est2.fit_transform(A_sparse)
- H1 = est1.components_
- H2 = est2.components_
+ W1 = est1.fit_transform(A)
+ W2 = est2.fit_transform(A_sparse)
+ H1 = est1.components_
+ H2 = est2.components_
- assert_array_almost_equal(W1, W2)
- assert_array_almost_equal(H1, H2)
+ assert_array_almost_equal(W1, W2)
+ assert_array_almost_equal(H1, H2)
-@ignore_warnings
def test_sparse_transform():
# Test that transform works on sparse data. Issue #2124
@@ -203,34 +165,29 @@ def test_sparse_transform():
A[A > 1.0] = 0
A = csc_matrix(A)
- for solver in ('pg', 'cd'):
- model = NMF(solver=solver, random_state=0, tol=1e-4, n_components=2)
- A_fit_tr = model.fit_transform(A)
- A_tr = model.transform(A)
- assert_array_almost_equal(A_fit_tr, A_tr, decimal=1)
+ model = NMF(random_state=0, tol=1e-4, n_components=2)
+ A_fit_tr = model.fit_transform(A)
+ A_tr = model.transform(A)
+ assert_array_almost_equal(A_fit_tr, A_tr, decimal=1)
-@ignore_warnings
def test_non_negative_factorization_consistency():
# Test that the function is called in the same way, either directly
# or through the NMF class
A = np.abs(random_state.randn(10, 10))
A[:, 2 * np.arange(5)] = 0
- for solver in ('pg', 'cd'):
- W_nmf, H, _ = non_negative_factorization(
- A, solver=solver, random_state=1, tol=1e-2)
- W_nmf_2, _, _ = non_negative_factorization(
- A, H=H, update_H=False, solver=solver, random_state=1, tol=1e-2)
+ W_nmf, H, _ = non_negative_factorization(A, random_state=1, tol=1e-2)
+ W_nmf_2, _, _ = non_negative_factorization(
+ A, H=H, update_H=False, random_state=1, tol=1e-2)
- model_class = NMF(solver=solver, random_state=1, tol=1e-2)
- W_cls = model_class.fit_transform(A)
- W_cls_2 = model_class.transform(A)
- assert_array_almost_equal(W_nmf, W_cls, decimal=10)
- assert_array_almost_equal(W_nmf_2, W_cls_2, decimal=10)
+ model_class = NMF(random_state=1, tol=1e-2)
+ W_cls = model_class.fit_transform(A)
+ W_cls_2 = model_class.transform(A)
+ assert_array_almost_equal(W_nmf, W_cls, decimal=10)
+ assert_array_almost_equal(W_nmf_2, W_cls_2, decimal=10)
-@ignore_warnings
def test_non_negative_factorization_checking():
A = np.ones((2, 2))
# Test parameters checking is public function
diff --git a/sklearn/discriminant_analysis.py b/sklearn/discriminant_analysis.py
index 628314a013..04180f2843 100644
--- a/sklearn/discriminant_analysis.py
+++ b/sklearn/discriminant_analysis.py
@@ -56,7 +56,8 @@ def _cov(X, shrinkage=None):
sc = StandardScaler() # standardize features
X = sc.fit_transform(X)
s = ledoit_wolf(X)[0]
- s = sc.scale_[:, np.newaxis] * s * sc.scale_[np.newaxis, :] # rescale
+ # rescale
+ s = sc.scale_[:, np.newaxis] * s * sc.scale_[np.newaxis, :]
elif shrinkage == 'empirical':
s = empirical_covariance(X)
else:
@@ -407,15 +408,15 @@ class LinearDiscriminantAnalysis(BaseEstimator, LinearClassifierMixin,
self.coef_ = np.dot(coef, self.scalings_.T)
self.intercept_ -= np.dot(self.xbar_, self.coef_.T)
- def fit(self, X, y, store_covariance=None, tol=None):
+ def fit(self, X, y):
"""Fit LinearDiscriminantAnalysis model according to the given
training data and parameters.
- .. versionchanged:: 0.17
- Deprecated *store_covariance* have been moved to main constructor.
+ .. versionchanged:: 0.19
+ *store_covariance* has been moved to main constructor.
- .. versionchanged:: 0.17
- Deprecated *tol* have been moved to main constructor.
+ .. versionchanged:: 0.19
+ *tol* has been moved to main constructor.
Parameters
----------
@@ -425,20 +426,6 @@ class LinearDiscriminantAnalysis(BaseEstimator, LinearClassifierMixin,
y : array, shape (n_samples,)
Target values.
"""
- if store_covariance:
- warnings.warn("The parameter 'store_covariance' is deprecated as "
- "of version 0.17 and will be removed in 0.19. The "
- "parameter is no longer necessary because the value "
- "is set via the estimator initialisation or "
- "set_params method.", DeprecationWarning)
- self.store_covariance = store_covariance
- if tol:
- warnings.warn("The parameter 'tol' is deprecated as of version "
- "0.17 and will be removed in 0.19. The parameter is "
- "no longer necessary because the value is set via "
- "the estimator initialisation or set_params method.",
- DeprecationWarning)
- self.tol = tol
X, y = check_X_y(X, y, ensure_min_samples=2, estimator=self)
self.classes_ = unique_labels(y)
@@ -630,14 +617,14 @@ class QuadraticDiscriminantAnalysis(BaseEstimator, ClassifierMixin):
self.store_covariances = store_covariances
self.tol = tol
- def fit(self, X, y, store_covariances=None, tol=None):
+ def fit(self, X, y):
"""Fit the model according to the given training data and parameters.
- .. versionchanged:: 0.17
- Deprecated *store_covariance* have been moved to main constructor.
+ .. versionchanged:: 0.19
+ *store_covariance* has been moved to main constructor.
- .. versionchanged:: 0.17
- Deprecated *tol* have been moved to main constructor.
+ .. versionchanged:: 0.19
+ *tol* has been moved to main constructor.
Parameters
----------
@@ -648,20 +635,6 @@ class QuadraticDiscriminantAnalysis(BaseEstimator, ClassifierMixin):
y : array, shape = [n_samples]
Target values (integers)
"""
- if store_covariances:
- warnings.warn("The parameter 'store_covariances' is deprecated as "
- "of version 0.17 and will be removed in 0.19. The "
- "parameter is no longer necessary because the value "
- "is set via the estimator initialisation or "
- "set_params method.", DeprecationWarning)
- self.store_covariances = store_covariances
- if tol:
- warnings.warn("The parameter 'tol' is deprecated as of version "
- "0.17 and will be removed in 0.19. The parameter is "
- "no longer necessary because the value is set via "
- "the estimator initialisation or set_params method.",
- DeprecationWarning)
- self.tol = tol
X, y = check_X_y(X, y)
check_classification_targets(y)
self.classes_, y = np.unique(y, return_inverse=True)
diff --git a/sklearn/ensemble/forest.py b/sklearn/ensemble/forest.py
index 3d252dfa4f..5ab0a0b191 100644
--- a/sklearn/ensemble/forest.py
+++ b/sklearn/ensemble/forest.py
@@ -53,7 +53,6 @@ from scipy.sparse import hstack as sparse_hstack
from ..base import ClassifierMixin, RegressorMixin
from ..externals.joblib import Parallel, delayed
from ..externals import six
-from ..feature_selection.from_model import _LearntSelectorMixin
from ..metrics import r2_score
from ..preprocessing import OneHotEncoder
from ..tree import (DecisionTreeClassifier, DecisionTreeRegressor,
@@ -124,8 +123,7 @@ def _parallel_build_trees(tree, forest, X, y, sample_weight, tree_idx, n_trees,
return tree
-class BaseForest(six.with_metaclass(ABCMeta, BaseEnsemble,
- _LearntSelectorMixin)):
+class BaseForest(six.with_metaclass(ABCMeta, BaseEnsemble)):
"""Base class for forests of trees.
Warning: This class should not be used directly. Use derived classes
@@ -473,17 +471,12 @@ class ForestClassifier(six.with_metaclass(ABCMeta, BaseForest,
y = y_store_unique_indices
if self.class_weight is not None:
- valid_presets = ('auto', 'balanced', 'subsample', 'balanced_subsample')
+ valid_presets = ('balanced', 'balanced_subsample')
if isinstance(self.class_weight, six.string_types):
if self.class_weight not in valid_presets:
raise ValueError('Valid presets for class_weight include '
'"balanced" and "balanced_subsample". Given "%s".'
% self.class_weight)
- if self.class_weight == "subsample":
- warn("class_weight='subsample' is deprecated in 0.17 and"
- "will be removed in 0.19. It was replaced by "
- "class_weight='balanced_subsample' using the balanced"
- "strategy.", DeprecationWarning)
if self.warm_start:
warn('class_weight presets "balanced" or "balanced_subsample" are '
'not recommended for warm_start if the fitted data '
@@ -495,19 +488,14 @@ class ForestClassifier(six.with_metaclass(ABCMeta, BaseForest,
'distributions. Pass the resulting weights as the '
'class_weight parameter.')
- if (self.class_weight not in ['subsample', 'balanced_subsample'] or
+ if (self.class_weight != 'balanced_subsample' or
not self.bootstrap):
- if self.class_weight == 'subsample':
- class_weight = 'auto'
- elif self.class_weight == "balanced_subsample":
+ if self.class_weight == "balanced_subsample":
class_weight = "balanced"
else:
class_weight = self.class_weight
- with warnings.catch_warnings():
- if class_weight == "auto":
- warnings.simplefilter('ignore', DeprecationWarning)
- expanded_class_weight = compute_sample_weight(class_weight,
- y_original)
+ expanded_class_weight = compute_sample_weight(class_weight,
+ y_original)
return y, expanded_class_weight
@@ -1685,9 +1673,7 @@ class RandomTreesEmbedding(BaseForest):
X_transformed : sparse matrix, shape=(n_samples, n_out)
Transformed dataset.
"""
- # ensure_2d=False because there are actually unit test checking we fail
- # for 1d.
- X = check_array(X, accept_sparse=['csc'], ensure_2d=False)
+ X = check_array(X, accept_sparse=['csc'])
if issparse(X):
# Pre-sort indices to avoid that each individual tree of the
# ensemble sorts the indices.
diff --git a/sklearn/ensemble/gradient_boosting.py b/sklearn/ensemble/gradient_boosting.py
index 2db5b574ad..9bef9635ea 100644
--- a/sklearn/ensemble/gradient_boosting.py
+++ b/sklearn/ensemble/gradient_boosting.py
@@ -31,7 +31,6 @@ from ..base import BaseEstimator
from ..base import ClassifierMixin
from ..base import RegressorMixin
from ..externals import six
-from ..feature_selection.from_model import _LearntSelectorMixin
from ._gradient_boosting import predict_stages
from ._gradient_boosting import predict_stage
@@ -55,7 +54,6 @@ from ..utils import check_array
from ..utils import check_X_y
from ..utils import column_or_1d
from ..utils import check_consistent_length
-from ..utils import deprecated
from ..utils.extmath import logsumexp
from ..utils.fixes import expit
from ..utils.fixes import bincount
@@ -715,8 +713,7 @@ class VerboseReporter(object):
self.verbose_mod *= 10
-class BaseGradientBoosting(six.with_metaclass(ABCMeta, BaseEnsemble,
- _LearntSelectorMixin)):
+class BaseGradientBoosting(six.with_metaclass(ABCMeta, BaseEnsemble)):
"""Abstract base class for Gradient Boosting. """
@abstractmethod
@@ -1125,30 +1122,6 @@ class BaseGradientBoosting(six.with_metaclass(ABCMeta, BaseEnsemble,
predict_stages(self.estimators_, X, self.learning_rate, score)
return score
- @deprecated(" and will be removed in 0.19")
- def decision_function(self, X):
- """Compute the decision function of ``X``.
-
- Parameters
- ----------
- X : array-like of shape = [n_samples, n_features]
- The input samples.
-
- Returns
- -------
- score : array, shape = [n_samples, n_classes] or [n_samples]
- The decision function of the input samples. The order of the
- classes corresponds to that in the attribute `classes_`.
- Regression and binary classification produce an array of shape
- [n_samples].
- """
-
- self._check_initialized()
- X = self.estimators_[0, 0]._validate_X_predict(X, check_input=True)
- score = self._decision_function(X)
- if score.shape[1] == 1:
- return score.ravel()
- return score
def _staged_decision_function(self, X):
"""Compute decision function of ``X`` for each iteration.
@@ -1177,30 +1150,6 @@ class BaseGradientBoosting(six.with_metaclass(ABCMeta, BaseEnsemble,
predict_stage(self.estimators_, i, X, self.learning_rate, score)
yield score.copy()
- @deprecated(" and will be removed in 0.19")
- def staged_decision_function(self, X):
- """Compute decision function of ``X`` for each iteration.
-
- This method allows monitoring (i.e. determine error on testing set)
- after each stage.
-
- Parameters
- ----------
- X : array-like of shape = [n_samples, n_features]
- The input samples.
-
- Returns
- -------
- score : generator of array, shape = [n_samples, k]
- The decision function of the input samples. The order of the
- classes corresponds to that in the attribute `classes_`.
- Regression and binary classification are special cases with
- ``k == 1``, otherwise ``k==n_classes``.
- """
- for dec in self._staged_decision_function(X):
- # no yield from in Python2.X
- yield dec
-
@property
def feature_importances_(self):
"""Return the feature importances (the higher, the more important the
diff --git a/sklearn/ensemble/iforest.py b/sklearn/ensemble/iforest.py
index 85b532db69..441d7078aa 100644
--- a/sklearn/ensemble/iforest.py
+++ b/sklearn/ensemble/iforest.py
@@ -154,9 +154,7 @@ class IsolationForest(BaseBagging):
self : object
Returns self.
"""
- # ensure_2d=False because there are actually unit test checking we fail
- # for 1d.
- X = check_array(X, accept_sparse=['csc'], ensure_2d=False)
+ X = check_array(X, accept_sparse=['csc'])
if issparse(X):
# Pre-sort indices to avoid that each individual tree of the
# ensemble sorts the indices.
diff --git a/sklearn/ensemble/tests/test_forest.py b/sklearn/ensemble/tests/test_forest.py
index 5ff4cf851f..9d7f796af6 100644
--- a/sklearn/ensemble/tests/test_forest.py
+++ b/sklearn/ensemble/tests/test_forest.py
@@ -208,12 +208,6 @@ def check_importances(name, criterion, X, y):
assert_equal(importances.shape[0], 10)
assert_equal(n_important, 3)
- # XXX: Remove this test in 0.19 after transform support to estimators
- # is removed.
- X_new = assert_warns(
- DeprecationWarning, est.transform, X, threshold="mean")
- assert_less(0 < X_new.shape[1], X.shape[1])
-
# Check with parallel
importances = est.feature_importances_
est.set_params(n_jobs=2)
@@ -968,11 +962,9 @@ def check_class_weight_balanced_and_bootstrap_multi_output(name):
clf = ForestClassifier(class_weight=[{-1: 0.5, 1: 1.}, {-2: 1., 2: 1.}],
random_state=0)
clf.fit(X, _y)
- # smoke test for subsample and balanced subsample
+ # smoke test for balanced subsample
clf = ForestClassifier(class_weight='balanced_subsample', random_state=0)
clf.fit(X, _y)
- clf = ForestClassifier(class_weight='subsample', random_state=0)
- ignore_warnings(clf.fit)(X, _y)
def test_class_weight_balanced_and_bootstrap_multi_output():
@@ -991,7 +983,7 @@ def check_class_weight_errors(name):
assert_raises(ValueError, clf.fit, X, _y)
# Warning warm_start with preset
- clf = ForestClassifier(class_weight='auto', warm_start=True,
+ clf = ForestClassifier(class_weight='balanced', warm_start=True,
random_state=0)
assert_warns(UserWarning, clf.fit, X, y)
assert_warns(UserWarning, clf.fit, X, _y)
diff --git a/sklearn/ensemble/tests/test_gradient_boosting.py b/sklearn/ensemble/tests/test_gradient_boosting.py
index 634bc259a1..817122338c 100644
--- a/sklearn/ensemble/tests/test_gradient_boosting.py
+++ b/sklearn/ensemble/tests/test_gradient_boosting.py
@@ -299,15 +299,6 @@ def test_feature_importances():
clf.fit(X, y)
assert_true(hasattr(clf, 'feature_importances_'))
- # XXX: Remove this test in 0.19 after transform support to estimators
- # is removed.
- X_new = assert_warns(
- DeprecationWarning, clf.transform, X, threshold="mean")
- assert_less(X_new.shape[1], X.shape[1])
- feature_mask = (
- clf.feature_importances_ > clf.feature_importances_.mean())
- assert_array_almost_equal(X_new, X[:, feature_mask])
-
def test_probability_log():
# Predict probabilities.
@@ -1073,6 +1064,7 @@ def check_sparse_input(EstimatorClass, X, X_sparse, y):
np.array(sparse.staged_decision_function(X_sparse)),
np.array(sparse.staged_decision_function(X)))
+
@skip_if_32bit
def test_sparse_input():
ests = (GradientBoostingClassifier, GradientBoostingRegressor)
diff --git a/sklearn/feature_selection/from_model.py b/sklearn/feature_selection/from_model.py
index 31a862b601..c0adcdcb6f 100644
--- a/sklearn/feature_selection/from_model.py
+++ b/sklearn/feature_selection/from_model.py
@@ -4,11 +4,9 @@
import numpy as np
from .base import SelectorMixin
-from ..base import TransformerMixin, BaseEstimator, clone
+from ..base import 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
from ..utils.fixes import norm
@@ -78,71 +76,6 @@ def _calculate_threshold(estimator, importances, 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.
diff --git a/sklearn/feature_selection/tests/test_from_model.py b/sklearn/feature_selection/tests/test_from_model.py
index 7f303cf8e2..6efb6f405b 100644
--- a/sklearn/feature_selection/tests/test_from_model.py
+++ b/sklearn/feature_selection/tests/test_from_model.py
@@ -1,5 +1,4 @@
import numpy as np
-import scipy.sparse as sp
from sklearn.utils.testing import assert_true
from sklearn.utils.testing import assert_equal
@@ -9,7 +8,6 @@ from sklearn.utils.testing import assert_array_almost_equal
from sklearn.utils.testing import assert_array_equal
from sklearn.utils.testing import assert_almost_equal
from sklearn.utils.testing import assert_raises
-from sklearn.utils.testing import assert_warns
from sklearn.utils.testing import skip_if_32bit
from sklearn import datasets
@@ -25,28 +23,6 @@ data, y = iris.data, iris.target
rng = np.random.RandomState(0)
-def test_transform_linear_model():
- for clf in (LogisticRegression(C=0.1),
- LinearSVC(C=0.01, dual=False),
- SGDClassifier(alpha=0.001, n_iter=50, shuffle=True,
- random_state=0)):
- for thresh in (None, ".09*mean", "1e-5 * median"):
- for func in (np.array, sp.csr_matrix):
- X = func(data)
- clf.set_params(penalty="l1")
- clf.fit(X, y)
- X_new = assert_warns(
- DeprecationWarning, clf.transform, X, thresh)
- if isinstance(clf, SGDClassifier):
- assert_true(X_new.shape[1] <= X.shape[1])
- else:
- assert_less(X_new.shape[1], X.shape[1])
- clf.set_params(penalty="l2")
- clf.fit(X_new, y)
- pred = clf.predict(X_new)
- assert_greater(np.mean(pred == y), 0.7)
-
-
def test_invalid_input():
clf = SGDClassifier(alpha=0.1, n_iter=10, shuffle=True, random_state=None)
for threshold in ["gobbledigook", ".5 * gobbledigook"]:
diff --git a/sklearn/grid_search.py b/sklearn/grid_search.py
index 6c0101a559..835ad92021 100644
--- a/sklearn/grid_search.py
+++ b/sklearn/grid_search.py
@@ -741,7 +741,7 @@ class GridSearchCV(BaseSearchCV):
... # doctest: +NORMALIZE_WHITESPACE +ELLIPSIS
GridSearchCV(cv=None, error_score=...,
estimator=SVC(C=1.0, cache_size=..., class_weight=..., coef0=...,
- decision_function_shape=None, degree=..., gamma=...,
+ decision_function_shape='ovr', degree=..., gamma=...,
kernel='rbf', max_iter=-1, probability=False,
random_state=None, shrinking=True, tol=...,
verbose=False),
diff --git a/sklearn/lda.py b/sklearn/lda.py
deleted file mode 100644
index 9c3959b6bc..0000000000
--- a/sklearn/lda.py
+++ /dev/null
@@ -1,20 +0,0 @@
-import warnings
-from .discriminant_analysis import LinearDiscriminantAnalysis as _LDA
-
-warnings.warn("lda.LDA has been moved to "
- "discriminant_analysis.LinearDiscriminantAnalysis "
- "in 0.17 and will be removed in 0.19", DeprecationWarning)
-
-
-class LDA(_LDA):
- """
- Alias for
- :class:`sklearn.discriminant_analysis.LinearDiscriminantAnalysis`.
-
- .. deprecated:: 0.17
- This class will be removed in 0.19.
- Use
- :class:`sklearn.discriminant_analysis.LinearDiscriminantAnalysis`
- instead.
- """
- pass
diff --git a/sklearn/linear_model/base.py b/sklearn/linear_model/base.py
index 1dbb1a4cc7..7ac614a1cd 100644
--- a/sklearn/linear_model/base.py
+++ b/sklearn/linear_model/base.py
@@ -229,22 +229,6 @@ class LinearModel(six.with_metaclass(ABCMeta, BaseEstimator)):
def fit(self, X, y):
"""Fit model."""
- @deprecated(" and will be removed in 0.19.")
- def decision_function(self, X):
- """Decision function of the linear model.
-
- Parameters
- ----------
- X : {array-like, sparse matrix}, shape = (n_samples, n_features)
- Samples.
-
- Returns
- -------
- C : array, shape = (n_samples,)
- Returns predicted values.
- """
- return self._decision_function(X)
-
def _decision_function(self, X):
check_is_fitted(self, "coef_")
@@ -478,12 +462,6 @@ class LinearRegression(LinearModel, RegressorMixin):
self.copy_X = copy_X
self.n_jobs = n_jobs
- @property
- @deprecated("``residues_`` is deprecated and will be removed in 0.19")
- def residues_(self):
- """Get the residues of the fitted model."""
- return self._residues
-
def fit(self, X, y, sample_weight=None):
"""
Fit linear model.
diff --git a/sklearn/linear_model/coordinate_descent.py b/sklearn/linear_model/coordinate_descent.py
index 5871c29738..de33fec9ab 100644
--- a/sklearn/linear_model/coordinate_descent.py
+++ b/sklearn/linear_model/coordinate_descent.py
@@ -15,7 +15,7 @@ from scipy import sparse
from .base import LinearModel, _pre_fit
from ..base import RegressorMixin
from .base import _preprocess_data
-from ..utils import check_array, check_X_y, deprecated
+from ..utils import check_array, check_X_y
from ..utils.validation import check_random_state
from ..model_selection import check_cv
from ..externals.joblib import Parallel, delayed
@@ -746,21 +746,6 @@ class ElasticNet(LinearModel, RegressorMixin):
""" sparse representation of the fitted ``coef_`` """
return sparse.csr_matrix(self.coef_)
- @deprecated(" and will be removed in 0.19")
- def decision_function(self, X):
- """Decision function of the linear model
-
- Parameters
- ----------
- X : numpy array or scipy.sparse matrix of shape (n_samples, n_features)
-
- Returns
- -------
- T : array, shape (n_samples,)
- The predicted decision function
- """
- return self._decision_function(X)
-
def _decision_function(self, X):
"""Decision function of the linear model
diff --git a/sklearn/linear_model/logistic.py b/sklearn/linear_model/logistic.py
index e792371383..ac4973f1df 100644
--- a/sklearn/linear_model/logistic.py
+++ b/sklearn/linear_model/logistic.py
@@ -17,7 +17,6 @@ from scipy import optimize, sparse
from .base import LinearClassifierMixin, SparseCoefMixin, BaseEstimator
from .sag import sag_solver
-from ..feature_selection.from_model import _LearntSelectorMixin
from ..preprocessing import LabelEncoder, LabelBinarizer
from ..svm.base import _fit_liblinear
from ..utils import check_array, check_consistent_length, compute_class_weight
@@ -445,7 +444,7 @@ def _check_solver_option(solver, multi_class, penalty, dual):
def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
max_iter=100, tol=1e-4, verbose=0,
- solver='lbfgs', coef=None, copy=False,
+ solver='lbfgs', coef=None,
class_weight=None, dual=False, penalty='l2',
intercept_scaling=1., multi_class='ovr',
random_state=None, check_input=True,
@@ -502,10 +501,6 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
Initialization value for coefficients of logistic regression.
Useless for liblinear solver.
- copy : bool, default False
- Whether or not to produce a copy of the data. A copy is not required
- anymore. This parameter is deprecated and will be removed in 0.19.
-
class_weight : dict or 'balanced', optional
Weights associated with classes in the form ``{class_label: weight}``.
If not given, all classes are supposed to have weight one.
@@ -579,21 +574,19 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
-----
You might get slightly different results with the solver liblinear than
with the others since this uses LIBLINEAR which penalizes the intercept.
- """
- if copy:
- warnings.warn("A copy is not required anymore. The 'copy' parameter "
- "is deprecated and will be removed in 0.19.",
- DeprecationWarning)
+ .. versionchanged:: 0.19
+ The "copy" parameter was removed.
+ """
if isinstance(Cs, numbers.Integral):
Cs = np.logspace(-4, 4, Cs)
_check_solver_option(solver, multi_class, penalty, dual)
# Preprocessing.
- if check_input or copy:
+ if check_input:
X = check_array(X, accept_sparse='csr', dtype=np.float64)
- y = check_array(y, ensure_2d=False, copy=copy, dtype=None)
+ y = check_array(y, ensure_2d=False, dtype=None)
check_consistent_length(X, y)
_, n_features = X.shape
classes = np.unique(y)
@@ -632,8 +625,7 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
y_bin[~mask] = -1.
# for compute_class_weight
- # 'auto' is deprecated and will be removed in 0.19
- if class_weight in ("auto", "balanced"):
+ if class_weight == "balanced":
class_weight_ = compute_class_weight(class_weight, mask_classes,
y_bin)
sample_weight *= class_weight_[le.fit_transform(y_bin)]
@@ -945,7 +937,7 @@ def _log_reg_scoring_path(X, y, train, test, pos_class=None, Cs=10,
class LogisticRegression(BaseEstimator, LinearClassifierMixin,
- _LearntSelectorMixin, SparseCoefMixin):
+ SparseCoefMixin):
"""Logistic Regression (aka logit, MaxEnt) classifier.
In the multiclass case, the training algorithm uses the one-vs-rest (OvR)
@@ -1011,8 +1003,7 @@ class LogisticRegression(BaseEstimator, LinearClassifierMixin,
through the fit method) if sample_weight is specified.
.. versionadded:: 0.17
- *class_weight='balanced'* instead of deprecated
- *class_weight='auto'*.
+ *class_weight='balanced'*
max_iter : int, default: 100
Useful only for the newton-cg, sag and lbfgs solvers.
@@ -1238,7 +1229,7 @@ class LogisticRegression(BaseEstimator, LinearClassifierMixin,
backend=backend)(
path_func(X, y, pos_class=class_, Cs=[self.C],
fit_intercept=self.fit_intercept, tol=self.tol,
- verbose=self.verbose, solver=self.solver, copy=False,
+ verbose=self.verbose, solver=self.solver,
multi_class=self.multi_class, max_iter=self.max_iter,
class_weight=self.class_weight, check_input=False,
random_state=self.random_state, coef=warm_start_coef_,
@@ -1313,7 +1304,7 @@ class LogisticRegression(BaseEstimator, LinearClassifierMixin,
class LogisticRegressionCV(LogisticRegression, BaseEstimator,
- LinearClassifierMixin, _LearntSelectorMixin):
+ LinearClassifierMixin):
"""Logistic Regression CV (aka logit, MaxEnt) classifier.
This class implements logistic regression using liblinear, newton-cg, sag
@@ -1559,11 +1550,6 @@ class LogisticRegressionCV(LogisticRegression, BaseEstimator,
check_classification_targets(y)
class_weight = self.class_weight
- if class_weight and not(isinstance(class_weight, dict) or
- class_weight in ['balanced', 'auto']):
- # 'auto' is deprecated and will be removed in 0.19
- raise ValueError("class_weight provided should be a "
- "dict or 'balanced'")
# Encode for string labels
label_encoder = LabelEncoder().fit(y)
@@ -1609,7 +1595,7 @@ class LogisticRegressionCV(LogisticRegression, BaseEstimator,
iter_classes = classes
# compute the class weights for the entire dataset y
- if class_weight in ("auto", "balanced"):
+ if class_weight == "balanced":
class_weight = compute_class_weight(class_weight,
np.arange(len(self.classes_)),
y)
@@ -1703,7 +1689,7 @@ class LogisticRegressionCV(LogisticRegression, BaseEstimator,
X, y, pos_class=encoded_label, Cs=[C_], solver=self.solver,
fit_intercept=self.fit_intercept, coef=coef_init,
max_iter=self.max_iter, tol=self.tol,
- penalty=self.penalty, copy=False,
+ penalty=self.penalty,
class_weight=class_weight,
multi_class=self.multi_class,
verbose=max(0, self.verbose - 1),
diff --git a/sklearn/linear_model/perceptron.py b/sklearn/linear_model/perceptron.py
index 76f8c648c7..d597181765 100644
--- a/sklearn/linear_model/perceptron.py
+++ b/sklearn/linear_model/perceptron.py
@@ -2,10 +2,9 @@
# License: BSD 3 clause
from .stochastic_gradient import BaseSGDClassifier
-from ..feature_selection.from_model import _LearntSelectorMixin
-class Perceptron(BaseSGDClassifier, _LearntSelectorMixin):
+class Perceptron(BaseSGDClassifier):
"""Perceptron
Read more in the :ref:`User Guide <perceptron>`.
diff --git a/sklearn/linear_model/stochastic_gradient.py b/sklearn/linear_model/stochastic_gradient.py
index 232ca90a77..f93ee0b4d7 100644
--- a/sklearn/linear_model/stochastic_gradient.py
+++ b/sklearn/linear_model/stochastic_gradient.py
@@ -13,9 +13,7 @@ from ..externals.joblib import Parallel, delayed
from .base import LinearClassifierMixin, SparseCoefMixin
from .base import make_dataset
from ..base import BaseEstimator, RegressorMixin
-from ..feature_selection.from_model import _LearntSelectorMixin
-from ..utils import (check_array, check_random_state, check_X_y,
- deprecated)
+from ..utils import check_array, check_random_state, check_X_y
from ..utils.extmath import safe_sparse_dot
from ..utils.multiclass import _check_partial_fit_first_call
from ..utils.validation import check_is_fitted
@@ -497,7 +495,7 @@ class BaseSGDClassifier(six.with_metaclass(ABCMeta, BaseSGD,
-------
self : returns an instance of self.
"""
- if self.class_weight in ['balanced', 'auto']:
+ if self.class_weight in ['balanced']:
raise ValueError("class_weight '{0}' is not supported for "
"partial_fit. In order to use 'balanced' weights,"
" use compute_class_weight('{0}', classes, y). "
@@ -545,7 +543,7 @@ class BaseSGDClassifier(six.with_metaclass(ABCMeta, BaseSGD,
sample_weight=sample_weight)
-class SGDClassifier(BaseSGDClassifier, _LearntSelectorMixin):
+class SGDClassifier(BaseSGDClassifier):
"""Linear classifiers (SVM, logistic regression, a.o.) with SGD training.
This estimator implements regularized linear models with stochastic
@@ -972,21 +970,6 @@ class BaseSGDRegressor(BaseSGD, RegressorMixin):
intercept_init=intercept_init,
sample_weight=sample_weight)
- @deprecated(" and will be removed in 0.19.")
- def decision_function(self, X):
- """Predict using the linear model
-
- Parameters
- ----------
- X : {array-like, sparse matrix}, shape (n_samples, n_features)
-
- Returns
- -------
- array, shape (n_samples,)
- Predicted target values per element in X.
- """
- return self._decision_function(X)
-
def _decision_function(self, X):
"""Predict using the linear model
@@ -1093,7 +1076,7 @@ class BaseSGDRegressor(BaseSGD, RegressorMixin):
self.intercept_ = np.atleast_1d(self.intercept_)
-class SGDRegressor(BaseSGDRegressor, _LearntSelectorMixin):
+class SGDRegressor(BaseSGDRegressor):
"""Linear model fitted by minimizing a regularized empirical loss with SGD
SGD stands for Stochastic Gradient Descent: the gradient of the loss is
diff --git a/sklearn/linear_model/tests/test_least_angle.py b/sklearn/linear_model/tests/test_least_angle.py
index 25cbf35032..fbd559695e 100644
--- a/sklearn/linear_model/tests/test_least_angle.py
+++ b/sklearn/linear_model/tests/test_least_angle.py
@@ -370,8 +370,6 @@ def test_multitarget():
for estimator in (linear_model.LassoLars(), linear_model.Lars()):
estimator.fit(X, Y)
Y_pred = estimator.predict(X)
- Y_dec = assert_warns(DeprecationWarning, estimator.decision_function, X)
- assert_array_almost_equal(Y_pred, Y_dec)
alphas, active, coef, path = (estimator.alphas_, estimator.active_,
estimator.coef_, estimator.coef_path_)
for k in range(n_targets):
diff --git a/sklearn/linear_model/tests/test_logistic.py b/sklearn/linear_model/tests/test_logistic.py
index 7fcee53651..ec2be517bf 100644
--- a/sklearn/linear_model/tests/test_logistic.py
+++ b/sklearn/linear_model/tests/test_logistic.py
@@ -10,7 +10,6 @@ from sklearn.utils.testing import assert_greater
from sklearn.utils.testing import assert_raises
from sklearn.utils.testing import assert_true
from sklearn.utils.testing import assert_warns
-from sklearn.utils.testing import assert_warns_message
from sklearn.utils.testing import raises
from sklearn.utils.testing import ignore_warnings
from sklearn.utils.testing import assert_raise_message
@@ -733,16 +732,6 @@ def test_logistic_regression_class_weights():
assert_array_almost_equal(clf1.coef_, clf2.coef_, decimal=6)
-def test_multinomial_logistic_regression_with_classweight_auto():
- X, y = iris.data, iris.target
- model = LogisticRegression(multi_class='multinomial',
- class_weight='auto', solver='lbfgs')
- # 'auto' is deprecated and will be removed in 0.19
- assert_warns_message(DeprecationWarning,
- "class_weight='auto' heuristic is deprecated",
- model.fit, X, y)
-
-
def test_logistic_regression_convergence_warnings():
# Test that warnings are raised if model does not converge
diff --git a/sklearn/metrics/base.py b/sklearn/metrics/base.py
index 0ad96c1afd..b8bbab3093 100644
--- a/sklearn/metrics/base.py
+++ b/sklearn/metrics/base.py
@@ -19,15 +19,6 @@ import numpy as np
from ..utils import check_array, check_consistent_length
from ..utils.multiclass import type_of_target
-from ..exceptions import UndefinedMetricWarning as _UndefinedMetricWarning
-from ..utils import deprecated
-
-
-@deprecated("UndefinedMetricWarning has been moved into the sklearn.exceptions"
- " module. It will not be available here from version 0.19")
-class UndefinedMetricWarning(_UndefinedMetricWarning):
- pass
-
def _average_binary_score(binary_metric, y_true, y_score, average,
sample_weight=None):
diff --git a/sklearn/metrics/regression.py b/sklearn/metrics/regression.py
index e4af5b2921..0450c39533 100644
--- a/sklearn/metrics/regression.py
+++ b/sklearn/metrics/regression.py
@@ -29,7 +29,6 @@ from ..utils.validation import check_array, check_consistent_length
from ..utils.validation import column_or_1d
from ..externals.six import string_types
-import warnings
__ALL__ = [
"mean_absolute_error",
@@ -436,9 +435,8 @@ def explained_variance_score(y_true, y_pred,
return np.average(output_scores, weights=avg_weights)
-def r2_score(y_true, y_pred,
- sample_weight=None,
- multioutput=None):
+def r2_score(y_true, y_pred, sample_weight=None,
+ multioutput="uniform_average"):
"""R^2 (coefficient of determination) regression score function.
Best possible score is 1.0 and it can be negative (because the
@@ -464,9 +462,7 @@ def r2_score(y_true, y_pred,
Defines aggregating of multiple output scores.
Array-like value defines weights used to average scores.
- Default value corresponds to 'variance_weighted', this behaviour is
- deprecated since version 0.17 and will be changed to 'uniform_average'
- starting from 0.19.
+ Default is "uniform_average".
'raw_values' :
Returns a full set of scores in case of multioutput input.
@@ -478,6 +474,9 @@ def r2_score(y_true, y_pred,
Scores of all outputs are averaged, weighted by the variances
of each individual output.
+ .. versionchanged:: 0.19
+ Default value of multioutput is 'uniform_average'.
+
Returns
-------
z : float or ndarray of floats
@@ -543,13 +542,6 @@ def r2_score(y_true, y_pred,
# arbitrary set to zero to avoid -inf scores, having a constant
# y_true is not interesting for scoring a regression anyway
output_scores[nonzero_numerator & ~nonzero_denominator] = 0.
- if multioutput is None and y_true.shape[1] != 1:
- warnings.warn("Default 'multioutput' behavior now corresponds to "
- "'variance_weighted' value which is deprecated since "
- "0.17, it will be changed to 'uniform_average' "
- "starting from 0.19.",
- DeprecationWarning)
- multioutput = 'variance_weighted'
if isinstance(multioutput, string_types):
if multioutput == 'raw_values':
# return scores individually
diff --git a/sklearn/model_selection/_search.py b/sklearn/model_selection/_search.py
index e1d744ceab..566ec8c996 100644
--- a/sklearn/model_selection/_search.py
+++ b/sklearn/model_selection/_search.py
@@ -804,7 +804,7 @@ class GridSearchCV(BaseSearchCV):
... # doctest: +NORMALIZE_WHITESPACE +ELLIPSIS
GridSearchCV(cv=None, error_score=...,
estimator=SVC(C=1.0, cache_size=..., class_weight=..., coef0=...,
- decision_function_shape=None, degree=..., gamma=...,
+ decision_function_shape='ovr', degree=..., gamma=...,
kernel='rbf', max_iter=-1, probability=False,
random_state=None, shrinking=True, tol=...,
verbose=False),
diff --git a/sklearn/pipeline.py b/sklearn/pipeline.py
index d8fa137d70..3b47eff30a 100644
--- a/sklearn/pipeline.py
+++ b/sklearn/pipeline.py
@@ -10,7 +10,6 @@ estimator, as a chain of transforms and estimators.
# License: BSD
from collections import defaultdict
-from warnings import warn
from abc import ABCMeta, abstractmethod
import numpy as np
@@ -470,10 +469,6 @@ class Pipeline(_BasePipeline):
return self._inverse_transform
def _inverse_transform(self, X):
- if hasattr(X, 'ndim') and X.ndim == 1:
- warn("From version 0.19, a 1d X will not be reshaped in"
- " pipeline.inverse_transform any more.", FutureWarning)
- X = X[None, :]
Xt = X
for name, transform in self.steps[::-1]:
if transform is not None:
diff --git a/sklearn/preprocessing/data.py b/sklearn/preprocessing/data.py
index 5e896a8ab1..ee160a1a8c 100644
--- a/sklearn/preprocessing/data.py
+++ b/sklearn/preprocessing/data.py
@@ -17,7 +17,6 @@ from scipy import sparse
from ..base import BaseEstimator, TransformerMixin
from ..externals import six
from ..utils import check_array
-from ..utils import deprecated
from ..utils.extmath import row_norms
from ..utils.extmath import _incremental_mean_and_var
from ..utils.fixes import bincount
@@ -51,13 +50,6 @@ __all__ = [
'minmax_scale',
]
-DEPRECATION_MSG_1D = (
- "Passing 1d arrays as data is deprecated in 0.17 and will "
- "raise ValueError in 0.19. Reshape your data either using "
- "X.reshape(-1, 1) if your data has a single feature or "
- "X.reshape(1, -1) if it contains a single sample."
-)
-
def _handle_zeros_in_scale(scale, copy=True):
''' Makes sure that whenever scale is zero, we handle it correctly.
@@ -226,19 +218,19 @@ class MinMaxScaler(BaseEstimator, TransformerMixin):
Per feature minimum seen in the data
.. versionadded:: 0.17
- *data_min_* instead of deprecated *data_min*.
+ *data_min_*
data_max_ : ndarray, shape (n_features,)
Per feature maximum seen in the data
.. versionadded:: 0.17
- *data_max_* instead of deprecated *data_max*.
+ *data_max_*
data_range_ : ndarray, shape (n_features,)
Per feature range ``(data_max_ - data_min_)`` seen in the data
.. versionadded:: 0.17
- *data_range_* instead of deprecated *data_range*.
+ *data_range_*
See also
--------
@@ -249,18 +241,6 @@ class MinMaxScaler(BaseEstimator, TransformerMixin):
self.feature_range = feature_range
self.copy = copy
- @property
- @deprecated("Attribute data_range will be removed in "
- "0.19. Use ``data_range_`` instead")
- def data_range(self):
- return self.data_range_
-
- @property
- @deprecated("Attribute data_min will be removed in "
- "0.19. Use ``data_min_`` instead")
- def data_min(self):
- return self.data_min_
-
def _reset(self):
"""Reset internal data-dependent state of the scaler, if necessary.
@@ -314,12 +294,9 @@ class MinMaxScaler(BaseEstimator, TransformerMixin):
raise TypeError("MinMaxScaler does no support sparse input. "
"You may consider to use MaxAbsScaler instead.")
- X = check_array(X, copy=self.copy, ensure_2d=False, warn_on_dtype=True,
+ X = check_array(X, copy=self.copy, warn_on_dtype=True,
estimator=self, dtype=FLOAT_DTYPES)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
-
data_min = np.min(X, axis=0)
data_max = np.max(X, axis=0)
@@ -351,9 +328,7 @@ class MinMaxScaler(BaseEstimator, TransformerMixin):
"""
check_is_fitted(self, 'scale_')
- X = check_array(X, copy=self.copy, ensure_2d=False, dtype=FLOAT_DTYPES)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ X = check_array(X, copy=self.copy, dtype=FLOAT_DTYPES)
X *= self.scale_
X += self.min_
@@ -369,9 +344,7 @@ class MinMaxScaler(BaseEstimator, TransformerMixin):
"""
check_is_fitted(self, 'scale_')
- X = check_array(X, copy=self.copy, ensure_2d=False, dtype=FLOAT_DTYPES)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ X = check_array(X, copy=self.copy, dtype=FLOAT_DTYPES)
X -= self.min_
X /= self.scale_
@@ -419,13 +392,7 @@ def minmax_scale(X, feature_range=(0, 1), axis=0, copy=True):
MinMaxScaler: Performs scaling to a given range using the``Transformer`` API
(e.g. as part of a preprocessing :class:`sklearn.pipeline.Pipeline`).
""" # noqa
- # To allow retro-compatibility, we handle here the case of 1D-input
- # From 0.17, 1D-input are deprecated in scaler objects
- # Although, we want to allow the users to keep calling this function
- # with 1D-input.
-
- # Cast input to array, as we need to check ndim. Prior to 0.17, that was
- # done inside the scaler object fit_transform.
+ # Unlike the scaler object, this function allows 1d input.
# If copy is required, it will be done inside the scaler object.
X = check_array(X, copy=False, ensure_2d=False, warn_on_dtype=True,
dtype=FLOAT_DTYPES)
@@ -497,7 +464,7 @@ class StandardScaler(BaseEstimator, TransformerMixin):
Per feature relative scaling of the data.
.. versionadded:: 0.17
- *scale_* is recommended instead of deprecated *std_*.
+ *scale_*
mean_ : array of floats with shape [n_features]
The mean value for each feature in the training set.
@@ -523,12 +490,6 @@ class StandardScaler(BaseEstimator, TransformerMixin):
self.with_std = with_std
self.copy = copy
- @property
- @deprecated("Attribute ``std_`` will be removed in 0.19. "
- "Use ``scale_`` instead")
- def std_(self):
- return self.scale_
-
def _reset(self):
"""Reset internal data-dependent state of the scaler, if necessary.
@@ -579,11 +540,7 @@ class StandardScaler(BaseEstimator, TransformerMixin):
y : Passthrough for ``Pipeline`` compatibility.
"""
X = check_array(X, accept_sparse=('csr', 'csc'), copy=self.copy,
- ensure_2d=False, warn_on_dtype=True,
- estimator=self, dtype=FLOAT_DTYPES)
-
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ warn_on_dtype=True, estimator=self, dtype=FLOAT_DTYPES)
# Even in the case of `with_mean=False`, we update the mean anyway
# This is needed for the incremental computation of the var
@@ -641,13 +598,9 @@ class StandardScaler(BaseEstimator, TransformerMixin):
check_is_fitted(self, 'scale_')
copy = copy if copy is not None else self.copy
- X = check_array(X, accept_sparse='csr', copy=copy,
- ensure_2d=False, warn_on_dtype=True,
+ X = check_array(X, accept_sparse='csr', copy=copy, warn_on_dtype=True,
estimator=self, dtype=FLOAT_DTYPES)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
-
if sparse.issparse(X):
if self.with_mean:
raise ValueError(
@@ -779,10 +732,7 @@ class MaxAbsScaler(BaseEstimator, TransformerMixin):
y : Passthrough for ``Pipeline`` compatibility.
"""
X = check_array(X, accept_sparse=('csr', 'csc'), copy=self.copy,
- ensure_2d=False, estimator=self, dtype=FLOAT_DTYPES)
-
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ estimator=self, dtype=FLOAT_DTYPES)
if sparse.issparse(X):
mins, maxs = min_max_axis(X, axis=0)
@@ -812,10 +762,7 @@ class MaxAbsScaler(BaseEstimator, TransformerMixin):
"""
check_is_fitted(self, 'scale_')
X = check_array(X, accept_sparse=('csr', 'csc'), copy=self.copy,
- ensure_2d=False, estimator=self, dtype=FLOAT_DTYPES)
-
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ estimator=self, dtype=FLOAT_DTYPES)
if sparse.issparse(X):
inplace_column_scale(X, 1.0 / self.scale_)
@@ -833,9 +780,7 @@ class MaxAbsScaler(BaseEstimator, TransformerMixin):
"""
check_is_fitted(self, 'scale_')
X = check_array(X, accept_sparse=('csr', 'csc'), copy=self.copy,
- ensure_2d=False, estimator=self, dtype=FLOAT_DTYPES)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ estimator=self, dtype=FLOAT_DTYPES)
if sparse.issparse(X):
inplace_column_scale(X, self.scale_)
@@ -868,13 +813,8 @@ def maxabs_scale(X, axis=0, copy=True):
MaxAbsScaler: Performs scaling to the [-1, 1] range using the``Transformer`` API
(e.g. as part of a preprocessing :class:`sklearn.pipeline.Pipeline`).
""" # noqa
- # To allow retro-compatibility, we handle here the case of 1D-input
- # From 0.17, 1D-input are deprecated in scaler objects
- # Although, we want to allow the users to keep calling this function
- # with 1D-input.
+ # Unlike the scaler object, this function allows 1d input.
- # Cast input to array, as we need to check ndim. Prior to 0.17, that was
- # done inside the scaler object fit_transform.
# If copy is required, it will be done inside the scaler object.
X = check_array(X, accept_sparse=('csr', 'csc'), copy=False,
ensure_2d=False, dtype=FLOAT_DTYPES)
@@ -980,10 +920,7 @@ class RobustScaler(BaseEstimator, TransformerMixin):
def _check_array(self, X, copy):
"""Makes sure centering is not enabled for sparse matrices."""
X = check_array(X, accept_sparse=('csr', 'csc'), copy=self.copy,
- ensure_2d=False, estimator=self, dtype=FLOAT_DTYPES)
-
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
+ estimator=self, dtype=FLOAT_DTYPES)
if sparse.issparse(X):
if self.with_centering:
@@ -1004,8 +941,6 @@ class RobustScaler(BaseEstimator, TransformerMixin):
if sparse.issparse(X):
raise TypeError("RobustScaler cannot be fitted on sparse inputs")
X = self._check_array(X, self.copy)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
if self.with_centering:
self.center_ = np.median(X, axis=0)
@@ -1033,8 +968,6 @@ class RobustScaler(BaseEstimator, TransformerMixin):
if self.with_scaling:
check_is_fitted(self, 'scale_')
X = self._check_array(X, self.copy)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
if sparse.issparse(X):
if self.with_scaling:
@@ -1059,8 +992,6 @@ class RobustScaler(BaseEstimator, TransformerMixin):
if self.with_scaling:
check_is_fitted(self, 'scale_')
X = self._check_array(X, self.copy)
- if X.ndim == 1:
- warnings.warn(DEPRECATION_MSG_1D, DeprecationWarning)
if sparse.issparse(X):
if self.with_scaling:
diff --git a/sklearn/preprocessing/tests/test_data.py b/sklearn/preprocessing/tests/test_data.py
index d76e008972..7a51049b60 100644
--- a/sklearn/preprocessing/tests/test_data.py
+++ b/sklearn/preprocessing/tests/test_data.py
@@ -27,7 +27,6 @@ from sklearn.utils.testing import assert_true
from sklearn.utils.testing import assert_false
from sklearn.utils.testing import assert_warns_message
from sklearn.utils.testing import assert_no_warnings
-from sklearn.utils.testing import ignore_warnings
from sklearn.utils.testing import assert_allclose
from sklearn.utils.testing import skip_if_32bit
@@ -790,12 +789,12 @@ def test_scale_sparse_with_mean_raise_exception():
def test_scale_input_finiteness_validation():
# Check if non finite inputs raise ValueError
- X = [np.nan, 5, 6, 7, 8]
+ X = [[np.nan, 5, 6, 7, 8]]
assert_raises_regex(ValueError,
"Input contains NaN, infinity or a value too large",
scale, X)
- X = [np.inf, 5, 6, 7, 8]
+ X = [[np.inf, 5, 6, 7, 8]]
assert_raises_regex(ValueError,
"Input contains NaN, infinity or a value too large",
scale, X)
@@ -1019,22 +1018,6 @@ def test_maxabs_scaler_transform_one_row_csr():
assert_array_almost_equal(X.toarray(), X_scaled_back.toarray())
-def test_deprecation_minmax_scaler():
- rng = np.random.RandomState(0)
- X = rng.random_sample((5, 4))
- scaler = MinMaxScaler().fit(X)
-
- depr_message = ("Attribute data_range will be removed in "
- "0.19. Use ``data_range_`` instead")
- assert_warns_message(DeprecationWarning, depr_message, getattr, scaler,
- "data_range")
-
- depr_message = ("Attribute data_min will be removed in "
- "0.19. Use ``data_min_`` instead")
- assert_warns_message(DeprecationWarning, depr_message, getattr, scaler,
- "data_min")
-
-
def test_warning_scaling_integers():
# Check warning when scaling integer data
X = np.array([[1, 2, 0],
@@ -1443,17 +1426,6 @@ def test_fit_transform():
assert_array_equal(X_transformed, X_transformed2)
-def test_deprecation_standard_scaler():
- rng = np.random.RandomState(0)
- X = rng.random_sample((5, 4))
- scaler = StandardScaler().fit(X)
- depr_message = ("Function std_ is deprecated; Attribute ``std_`` will be "
- "removed in 0.19. Use ``scale_`` instead")
- std_ = assert_warns_message(DeprecationWarning, depr_message, getattr,
- scaler, "std_")
- assert_array_equal(std_, scaler.scale_)
-
-
def test_add_dummy_feature():
X = [[1, 0], [0, 1], [0, 1]]
X = add_dummy_feature(X)
diff --git a/sklearn/qda.py b/sklearn/qda.py
deleted file mode 100644
index 604d6a919d..0000000000
--- a/sklearn/qda.py
+++ /dev/null
@@ -1,20 +0,0 @@
-import warnings
-from .discriminant_analysis import QuadraticDiscriminantAnalysis as _QDA
-
-warnings.warn("qda.QDA has been moved to "
- "discriminant_analysis.QuadraticDiscriminantAnalysis "
- "in 0.17 and will be removed in 0.19.", DeprecationWarning)
-
-
-class QDA(_QDA):
- """
- Alias for
- :class:`sklearn.discriminant_analysis.QuadraticDiscriminantAnalysis`.
-
- .. deprecated:: 0.17
- This class will be removed in 0.19.
- Use
- :class:`sklearn.discriminant_analysis.QuadraticDiscriminantAnalysis`
- instead.
- """
- pass
diff --git a/sklearn/svm/base.py b/sklearn/svm/base.py
index b00130127f..3e416b0821 100644
--- a/sklearn/svm/base.py
+++ b/sklearn/svm/base.py
@@ -12,12 +12,11 @@ from ..preprocessing import LabelEncoder
from ..utils.multiclass import _ovr_decision_function
from ..utils import check_array, check_consistent_length, check_random_state
from ..utils import column_or_1d, check_X_y
-from ..utils import compute_class_weight, deprecated
+from ..utils import compute_class_weight
from ..utils.extmath import safe_sparse_dot
from ..utils.validation import check_is_fitted
from ..utils.multiclass import check_classification_targets
from ..externals import six
-from ..exceptions import ChangedBehaviorWarning
from ..exceptions import ConvergenceWarning
from ..exceptions import NotFittedError
@@ -368,24 +367,6 @@ class BaseLibSVM(six.with_metaclass(ABCMeta, BaseEstimator)):
X = np.asarray(kernel, dtype=np.float64, order='C')
return X
- @deprecated(" and will be removed in 0.19")
- def decision_function(self, X):
- """Distance of the samples X to the separating hyperplane.
-
- Parameters
- ----------
- X : array-like, shape (n_samples, n_features)
- For kernel="precomputed", the expected shape of X is
- [n_samples_test, n_samples_train].
-
- Returns
- -------
- X : array-like, shape (n_samples, n_class * (n_class-1) / 2)
- Returns the decision function of the sample for each class
- in the model.
- """
- return self._decision_function(X)
-
def _decision_function(self, X):
"""Distance of the samples X to the separating hyperplane.
@@ -545,11 +526,6 @@ class BaseSVC(six.with_metaclass(ABCMeta, BaseLibSVM, ClassifierMixin)):
n_classes)
"""
dec = self._decision_function(X)
- if self.decision_function_shape is None and len(self.classes_) > 2:
- warnings.warn("The decision_function_shape default value will "
- "change from 'ovo' to 'ovr' in 0.19. This will change "
- "the shape of the decision function returned by "
- "SVC.", ChangedBehaviorWarning)
if self.decision_function_shape == 'ovr' and len(self.classes_) > 2:
return _ovr_decision_function(dec < 0, -dec, len(self.classes_))
return dec
diff --git a/sklearn/svm/classes.py b/sklearn/svm/classes.py
index 1d269a02c9..7e920011d0 100644
--- a/sklearn/svm/classes.py
+++ b/sklearn/svm/classes.py
@@ -5,14 +5,13 @@ from .base import _fit_liblinear, BaseSVC, BaseLibSVM
from ..base import BaseEstimator, RegressorMixin
from ..linear_model.base import LinearClassifierMixin, SparseCoefMixin, \
LinearModel
-from ..feature_selection.from_model import _LearntSelectorMixin
from ..utils import check_X_y
from ..utils.validation import _num_samples
from ..utils.multiclass import check_classification_targets
class LinearSVC(BaseEstimator, LinearClassifierMixin,
- _LearntSelectorMixin, SparseCoefMixin):
+ SparseCoefMixin):
"""Linear Support Vector Classification.
Similar to SVC with parameter kernel='linear', but implemented in terms of
@@ -50,13 +49,13 @@ class LinearSVC(BaseEstimator, LinearClassifierMixin,
multi_class : string, 'ovr' or 'crammer_singer' (default='ovr')
Determines the multi-class strategy if `y` contains more than
two classes.
- ``"ovr"`` trains n_classes one-vs-rest classifiers, while ``"crammer_singer"``
- optimizes a joint objective over all classes.
+ ``"ovr"`` trains n_classes one-vs-rest classifiers, while
+ ``"crammer_singer"`` optimizes a joint objective over all classes.
While `crammer_singer` is interesting from a theoretical perspective
as it is consistent, it is seldom used in practice as it rarely leads
to better accuracy and is more expensive to compute.
- If ``"crammer_singer"`` is chosen, the options loss, penalty and dual will
- be ignored.
+ If ``"crammer_singer"`` is chosen, the options loss, penalty and dual
+ will be ignored.
fit_intercept : boolean, optional (default=True)
Whether to calculate the intercept for this model. If set
@@ -454,14 +453,14 @@ class SVC(BaseSVC):
max_iter : int, optional (default=-1)
Hard limit on iterations within solver, or -1 for no limit.
- decision_function_shape : 'ovo', 'ovr' or None, default=None
+ decision_function_shape : 'ovo', 'ovr', default='ovr'
Whether to return a one-vs-rest ('ovr') decision function of shape
(n_samples, n_classes) as all other classifiers, or the original
one-vs-one ('ovo') decision function of libsvm which has shape
(n_samples, n_classes * (n_classes - 1) / 2).
- The default of None will currently behave as 'ovo' for backward
- compatibility and raise a deprecation warning, but will change 'ovr'
- in 0.19.
+
+ .. versionchanged:: 0.19
+ decision_function_shape is 'ovr' by default.
.. versionadded:: 0.17
*decision_function_shape='ovr'* is recommended.
@@ -510,7 +509,7 @@ class SVC(BaseSVC):
>>> clf = SVC()
>>> clf.fit(X, y) #doctest: +NORMALIZE_WHITESPACE
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
>>> print(clf.predict([[-0.8, -1]]))
@@ -531,7 +530,7 @@ class SVC(BaseSVC):
def __init__(self, C=1.0, kernel='rbf', degree=3, gamma='auto',
coef0=0.0, shrinking=True, probability=False,
tol=1e-3, cache_size=200, class_weight=None,
- verbose=False, max_iter=-1, decision_function_shape=None,
+ verbose=False, max_iter=-1, decision_function_shape='ovr',
random_state=None):
super(SVC, self).__init__(
@@ -595,8 +594,8 @@ class NuSVC(BaseSVC):
class_weight : {dict, 'balanced'}, optional
Set the parameter C of class i to class_weight[i]*C for
SVC. If not given, all classes are supposed to have
- weight one. The "balanced" mode uses the values of y to automatically adjust
- weights inversely proportional to class frequencies as
+ weight one. The "balanced" mode uses the values of y to automatically
+ adjust weights inversely proportional to class frequencies as
``n_samples / (n_classes * np.bincount(y))``
verbose : bool, default: False
@@ -607,14 +606,14 @@ class NuSVC(BaseSVC):
max_iter : int, optional (default=-1)
Hard limit on iterations within solver, or -1 for no limit.
- decision_function_shape : 'ovo', 'ovr' or None, default=None
+ decision_function_shape : 'ovo', 'ovr', default='ovr'
Whether to return a one-vs-rest ('ovr') decision function of shape
(n_samples, n_classes) as all other classifiers, or the original
one-vs-one ('ovo') decision function of libsvm which has shape
(n_samples, n_classes * (n_classes - 1) / 2).
- The default of None will currently behave as 'ovo' for backward
- compatibility and raise a deprecation warning, but will change 'ovr'
- in 0.19.
+
+ .. versionchanged:: 0.19
+ decision_function_shape is 'ovr' by default.
.. versionadded:: 0.17
*decision_function_shape='ovr'* is recommended.
@@ -663,7 +662,7 @@ class NuSVC(BaseSVC):
>>> clf = NuSVC()
>>> clf.fit(X, y) #doctest: +NORMALIZE_WHITESPACE
NuSVC(cache_size=200, class_weight=None, coef0=0.0,
- decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',
+ decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
max_iter=-1, nu=0.5, probability=False, random_state=None,
shrinking=True, tol=0.001, verbose=False)
>>> print(clf.predict([[-0.8, -1]]))
@@ -679,10 +678,10 @@ class NuSVC(BaseSVC):
liblinear.
"""
- def __init__(self, nu=0.5, kernel='rbf', degree=3, gamma='auto',
- coef0=0.0, shrinking=True, probability=False,
- tol=1e-3, cache_size=200, class_weight=None, verbose=False,
- max_iter=-1, decision_function_shape=None, random_state=None):
+ def __init__(self, nu=0.5, kernel='rbf', degree=3, gamma='auto', coef0=0.0,
+ shrinking=True, probability=False, tol=1e-3, cache_size=200,
+ class_weight=None, verbose=False, max_iter=-1,
+ decision_function_shape='ovr', random_state=None):
super(NuSVC, self).__init__(
impl='nu_svc', kernel=kernel, degree=degree, gamma=gamma,
diff --git a/sklearn/svm/tests/test_svm.py b/sklearn/svm/tests/test_svm.py
index 88c4b04dfc..ce122a4fcf 100644
--- a/sklearn/svm/tests/test_svm.py
+++ b/sklearn/svm/tests/test_svm.py
@@ -20,7 +20,6 @@ from sklearn.utils.testing import assert_greater, assert_in, assert_less
from sklearn.utils.testing import assert_raises_regexp, assert_warns
from sklearn.utils.testing import assert_warns_message, assert_raise_message
from sklearn.utils.testing import ignore_warnings, assert_raises
-from sklearn.exceptions import ChangedBehaviorWarning
from sklearn.exceptions import ConvergenceWarning
from sklearn.exceptions import NotFittedError
from sklearn.multiclass import OneVsRestClassifier
@@ -87,17 +86,6 @@ def test_libsvm_iris():
assert_array_equal(pred, pred2)
-@ignore_warnings
-def test_single_sample_1d():
- # Test whether SVCs work on a single sample given as a 1-d array
-
- clf = svm.SVC().fit(X, Y)
- clf.predict(X[0])
-
- clf = svm.LinearSVC(random_state=0).fit(X, Y)
- clf.predict(X[0])
-
-
def test_precomputed():
# SVC with a precomputed kernel.
# We test it with a toy dataset and with iris.
@@ -382,13 +370,6 @@ def test_decision_function_shape():
dec = clf.decision_function(X_train)
assert_equal(dec.shape, (len(X_train), 10))
- # check deprecation warning
- clf = svm.SVC(kernel='linear', C=0.1).fit(X_train, y_train)
- msg = "change the shape of the decision function"
- dec = assert_warns_message(ChangedBehaviorWarning, msg,
- clf.decision_function, X_train)
- assert_equal(dec.shape, (len(X_train), 10))
-
def test_svr_predict():
# Test SVR's decision_function
diff --git a/sklearn/tests/test_discriminant_analysis.py b/sklearn/tests/test_discriminant_analysis.py
index d1049fa6de..a7a878a731 100644
--- a/sklearn/tests/test_discriminant_analysis.py
+++ b/sklearn/tests/test_discriminant_analysis.py
@@ -1,4 +1,3 @@
-import sys
import numpy as np
from sklearn.utils.testing import assert_array_equal
@@ -11,7 +10,6 @@ from sklearn.utils.testing import assert_raise_message
from sklearn.utils.testing import assert_warns
from sklearn.utils.testing import assert_greater
from sklearn.utils.testing import ignore_warnings
-from sklearn.utils.testing import SkipTest
from sklearn.datasets import make_blobs
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
@@ -19,16 +17,6 @@ from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis
from sklearn.discriminant_analysis import _cov
-# import reload
-version = sys.version_info
-if version[0] == 3:
- # Python 3+ import for reload. Builtin in Python2
- if version[1] == 3:
- reload = None
- else:
- from importlib import reload
-
-
# Data is just 6 separable points in the plane
X = np.array([[-2, -1], [-1, -1], [-1, -2], [1, 1], [1, 2], [2, 1]], dtype='f')
y = np.array([1, 1, 1, 2, 2, 2])
@@ -317,31 +305,6 @@ def test_qda_regularization():
assert_array_equal(y_pred5, y5)
-def test_deprecated_lda_qda_deprecation():
- if reload is None:
- raise SkipTest("Can't reload module on Python3.3")
-
- def import_lda_module():
- import sklearn.lda
- # ensure that we trigger DeprecationWarning even if the sklearn.lda
- # was loaded previously by another test.
- reload(sklearn.lda)
- return sklearn.lda
-
- lda = assert_warns(DeprecationWarning, import_lda_module)
- assert isinstance(lda.LDA(), LinearDiscriminantAnalysis)
-
- def import_qda_module():
- import sklearn.qda
- # ensure that we trigger DeprecationWarning even if the sklearn.qda
- # was loaded previously by another test.
- reload(sklearn.qda)
- return sklearn.qda
-
- qda = assert_warns(DeprecationWarning, import_qda_module)
- assert isinstance(qda.QDA(), QuadraticDiscriminantAnalysis)
-
-
def test_covariance():
x, y = make_blobs(n_samples=100, n_features=5,
centers=1, random_state=42)
diff --git a/sklearn/tests/test_multiclass.py b/sklearn/tests/test_multiclass.py
index 5d4140b5d7..b62e78e87c 100644
--- a/sklearn/tests/test_multiclass.py
+++ b/sklearn/tests/test_multiclass.py
@@ -334,7 +334,6 @@ def test_ovr_multilabel_predict_proba():
# Decision function only estimator.
decision_only = OneVsRestClassifier(svm.SVR()).fit(X_train, Y_train)
assert_false(hasattr(decision_only, 'predict_proba'))
- assert_true(hasattr(decision_only, 'decision_function'))
# Estimator with predict_proba disabled, depending on parameters.
decision_only = OneVsRestClassifier(svm.SVC(probability=False))
@@ -370,7 +369,6 @@ def test_ovr_single_label_predict_proba():
# Decision function only estimator.
decision_only = OneVsRestClassifier(svm.SVR()).fit(X_train, Y_train)
assert_false(hasattr(decision_only, 'predict_proba'))
- assert_true(hasattr(decision_only, 'decision_function'))
Y_pred = clf.predict(X_test)
Y_proba = clf.predict_proba(X_test)
diff --git a/sklearn/tests/test_pipeline.py b/sklearn/tests/test_pipeline.py
index 5f1f62cdce..fb105abd78 100644
--- a/sklearn/tests/test_pipeline.py
+++ b/sklearn/tests/test_pipeline.py
@@ -13,7 +13,6 @@ from sklearn.utils.testing import assert_false
from sklearn.utils.testing import assert_true
from sklearn.utils.testing import assert_array_equal
from sklearn.utils.testing import assert_array_almost_equal
-from sklearn.utils.testing import assert_warns_message
from sklearn.utils.testing import assert_dict_equal
from sklearn.base import clone, BaseEstimator
@@ -700,14 +699,6 @@ def test_classes_property():
assert_array_equal(clf.classes_, np.unique(y))
-def test_X1d_inverse_transform():
- transformer = Transf()
- pipeline = make_pipeline(transformer)
- X = np.ones(10)
- msg = "1d X will not be reshaped in pipeline.inverse_transform"
- assert_warns_message(FutureWarning, msg, pipeline.inverse_transform, X)
-
-
def test_set_feature_union_steps():
mult2 = Mult(2)
mult2.get_feature_names = lambda: ['x2']
diff --git a/sklearn/tree/tests/test_tree.py b/sklearn/tree/tests/test_tree.py
index c3e8e795b3..ff662e9af4 100644
--- a/sklearn/tree/tests/test_tree.py
+++ b/sklearn/tree/tests/test_tree.py
@@ -28,7 +28,6 @@ from sklearn.utils.testing import assert_greater_equal
from sklearn.utils.testing import assert_less
from sklearn.utils.testing import assert_less_equal
from sklearn.utils.testing import assert_true
-from sklearn.utils.testing import assert_warns
from sklearn.utils.testing import raises
from sklearn.utils.testing import ignore_warnings
@@ -382,11 +381,6 @@ def test_importances():
assert_equal(importances.shape[0], 10, "Failed with {0}".format(name))
assert_equal(n_important, 3, "Failed with {0}".format(name))
- X_new = assert_warns(
- DeprecationWarning, clf.transform, X, threshold="mean")
- assert_less(0, X_new.shape[1], "Failed with {0}".format(name))
- assert_less(X_new.shape[1], X.shape[1], "Failed with {0}".format(name))
-
# Check on iris that importances are the same for all builders
clf = DecisionTreeClassifier(random_state=0)
clf.fit(iris.data, iris.target)
@@ -529,7 +523,8 @@ def test_error():
X, y)
assert_raises(ValueError, TreeEstimator(max_depth=-1).fit, X, y)
assert_raises(ValueError, TreeEstimator(max_features=42).fit, X, y)
- assert_raises(ValueError, TreeEstimator(min_impurity_split=-1.0).fit, X, y)
+ assert_raises(ValueError, TreeEstimator(min_impurity_split=-1.0).fit,
+ X, y)
# Wrong dimensions
est = TreeEstimator()
@@ -602,7 +597,6 @@ def test_min_samples_split():
"Failed with {0}".format(name))
-
def test_min_samples_leaf():
# Test if leaves contain more than leaf_count training examples
X = np.asfortranarray(iris.data.astype(tree._tree.DTYPE))
@@ -867,7 +861,6 @@ def test_pickle():
"pickling with {1}".format(attribute, name))
-
def test_multioutput():
# Check estimators on multi-output problems.
X = [[-2, -1],
@@ -1287,18 +1280,19 @@ def check_sparse_input(tree, dataset, max_depth=None):
def test_sparse_input():
- for tree, dataset in product(SPARSE_TREES,
- ("clf_small", "toy", "digits", "multilabel",
- "sparse-pos", "sparse-neg", "sparse-mix",
- "zeros")):
+ for tree_type, dataset in product(SPARSE_TREES, ("clf_small", "toy",
+ "digits", "multilabel",
+ "sparse-pos",
+ "sparse-neg",
+ "sparse-mix", "zeros")):
max_depth = 3 if dataset == "digits" else None
- yield (check_sparse_input, tree, dataset, max_depth)
+ yield (check_sparse_input, tree_type, dataset, max_depth)
# Due to numerical instability of MSE and too strict test, we limit the
# maximal depth
- for tree, dataset in product(REG_TREES, ["boston", "reg_small"]):
- if tree in SPARSE_TREES:
- yield (check_sparse_input, tree, dataset, 2)
+ for tree_type, dataset in product(SPARSE_TREES, ["boston", "reg_small"]):
+ if tree_type in REG_TREES:
+ yield (check_sparse_input, tree_type, dataset, 2)
def check_sparse_parameters(tree, dataset):
@@ -1346,10 +1340,10 @@ def check_sparse_parameters(tree, dataset):
def test_sparse_parameters():
- for tree, dataset in product(SPARSE_TREES,
- ["sparse-pos", "sparse-neg", "sparse-mix",
- "zeros"]):
- yield (check_sparse_parameters, tree, dataset)
+ for tree_type, dataset in product(SPARSE_TREES, ["sparse-pos",
+ "sparse-neg",
+ "sparse-mix", "zeros"]):
+ yield (check_sparse_parameters, tree_type, dataset)
def check_sparse_criterion(tree, dataset):
@@ -1373,10 +1367,10 @@ def check_sparse_criterion(tree, dataset):
def test_sparse_criterion():
- for tree, dataset in product(SPARSE_TREES,
- ["sparse-pos", "sparse-neg", "sparse-mix",
- "zeros"]):
- yield (check_sparse_criterion, tree, dataset)
+ for tree_type, dataset in product(SPARSE_TREES, ["sparse-pos",
+ "sparse-neg",
+ "sparse-mix", "zeros"]):
+ yield (check_sparse_criterion, tree_type, dataset)
def check_explicit_sparse_zeros(tree, max_depth=3,
@@ -1449,8 +1443,8 @@ def check_explicit_sparse_zeros(tree, max_depth=3,
def test_explicit_sparse_zeros():
- for tree in SPARSE_TREES:
- yield (check_explicit_sparse_zeros, tree)
+ for tree_type in SPARSE_TREES:
+ yield (check_explicit_sparse_zeros, tree_type)
@ignore_warnings
diff --git a/sklearn/tree/tree.py b/sklearn/tree/tree.py
index c3567e864c..a4b6a2aac9 100644
--- a/sklearn/tree/tree.py
+++ b/sklearn/tree/tree.py
@@ -29,7 +29,6 @@ from ..base import BaseEstimator
from ..base import ClassifierMixin
from ..base import RegressorMixin
from ..externals import six
-from ..feature_selection.from_model import _LearntSelectorMixin
from ..utils import check_array
from ..utils import check_random_state
from ..utils import compute_sample_weight
@@ -71,8 +70,7 @@ SPARSE_SPLITTERS = {"best": _splitter.BestSparseSplitter,
# =============================================================================
-class BaseDecisionTree(six.with_metaclass(ABCMeta, BaseEstimator,
- _LearntSelectorMixin)):
+class BaseDecisionTree(six.with_metaclass(ABCMeta, BaseEstimator)):
"""Base class for decision trees.
Warning: This class should not be used directly.
@@ -739,7 +737,6 @@ class DecisionTreeClassifier(BaseDecisionTree, ClassifierMixin):
X_idx_sorted=X_idx_sorted)
return self
-
def predict_proba(self, X, check_input=True):
"""Predict class probabilities of the input samples X.
diff --git a/sklearn/utils/__init__.py b/sklearn/utils/__init__.py
index ac16ef9ad6..ede3bb45ce 100644
--- a/sklearn/utils/__init__.py
+++ b/sklearn/utils/__init__.py
@@ -13,17 +13,10 @@ from .validation import (as_float_array,
check_random_state, column_or_1d, check_array,
check_consistent_length, check_X_y, indexable,
check_symmetric)
-from .deprecation import deprecated
from .class_weight import compute_class_weight, compute_sample_weight
from ..externals.joblib import cpu_count
-from ..exceptions import ConvergenceWarning as _ConvergenceWarning
from ..exceptions import DataConversionWarning
-
-
-@deprecated("ConvergenceWarning has been moved into the sklearn.exceptions "
- "module. It will not be available here from version 0.19")
-class ConvergenceWarning(_ConvergenceWarning):
- pass
+from .deprecation import deprecated
__all__ = ["murmurhash3_32", "as_float_array",
@@ -32,7 +25,7 @@ __all__ = ["murmurhash3_32", "as_float_array",
"compute_class_weight", "compute_sample_weight",
"column_or_1d", "safe_indexing",
"check_consistent_length", "check_X_y", 'indexable',
- "check_symmetric", "indices_to_mask"]
+ "check_symmetric", "indices_to_mask", "deprecated"]
def safe_mask(X, mask):
diff --git a/sklearn/utils/class_weight.py b/sklearn/utils/class_weight.py
index 5b778423f1..119e5eabe9 100644
--- a/sklearn/utils/class_weight.py
+++ b/sklearn/utils/class_weight.py
@@ -2,7 +2,6 @@
# Manoj Kumar
# License: BSD 3 clause
-import warnings
import numpy as np
from ..externals import six
from ..utils.fixes import in1d
@@ -48,25 +47,16 @@ def compute_class_weight(class_weight, classes, y):
if class_weight is None or len(class_weight) == 0:
# uniform class weights
weight = np.ones(classes.shape[0], dtype=np.float64, order='C')
- elif class_weight in ['auto', 'balanced']:
+ elif class_weight == 'balanced':
# Find the weight of each class as present in y.
le = LabelEncoder()
y_ind = le.fit_transform(y)
if not all(np.in1d(classes, le.classes_)):
raise ValueError("classes should have valid labels that are in y")
- # inversely proportional to the number of samples in the class
- if class_weight == 'auto':
- recip_freq = 1. / bincount(y_ind)
- weight = recip_freq[le.transform(classes)] / np.mean(recip_freq)
- warnings.warn("The class_weight='auto' heuristic is deprecated in"
- " 0.17 in favor of a new heuristic "
- "class_weight='balanced'. 'auto' will be removed in"
- " 0.19", DeprecationWarning)
- else:
- recip_freq = len(y) / (len(le.classes_) *
- bincount(y_ind).astype(np.float64))
- weight = recip_freq[le.transform(classes)]
+ recip_freq = len(y) / (len(le.classes_) *
+ bincount(y_ind).astype(np.float64))
+ weight = recip_freq[le.transform(classes)]
else:
# user-defined dictionary
weight = np.ones(classes.shape[0], dtype=np.float64, order='C')
@@ -107,8 +97,8 @@ def compute_sample_weight(class_weight, y, indices=None):
Array of indices to be used in a subsample. Can be of length less than
n_samples in the case of a subsample, or equal to n_samples in the
case of a bootstrap subsample with repeated indices. If None, the
- sample weight will be calculated over the full sample. Only "auto" is
- supported for class_weight if this is provided.
+ sample weight will be calculated over the full sample. Only "balanced"
+ is supported for class_weight if this is provided.
Returns
-------
@@ -122,7 +112,7 @@ def compute_sample_weight(class_weight, y, indices=None):
n_outputs = y.shape[1]
if isinstance(class_weight, six.string_types):
- if class_weight not in ['balanced', 'auto']:
+ if class_weight not in ['balanced']:
raise ValueError('The only valid preset for class_weight is '
'"balanced". Given "%s".' % class_weight)
elif (indices is not None and
@@ -145,7 +135,7 @@ def compute_sample_weight(class_weight, y, indices=None):
classes_full = np.unique(y_full)
classes_missing = None
- if class_weight in ['balanced', 'auto'] or n_outputs == 1:
+ if class_weight == 'balanced' or n_outputs == 1:
class_weight_k = class_weight
else:
class_weight_k = class_weight[k]
diff --git a/sklearn/utils/estimator_checks.py b/sklearn/utils/estimator_checks.py
index 1af836a2a1..db5569af28 100644
--- a/sklearn/utils/estimator_checks.py
+++ b/sklearn/utils/estimator_checks.py
@@ -42,7 +42,6 @@ from sklearn.random_projection import BaseRandomProjection
from sklearn.feature_selection import SelectKBest
from sklearn.svm.base import BaseLibSVM
from sklearn.pipeline import make_pipeline
-from sklearn.decomposition import NMF, ProjectedGradientNMF
from sklearn.exceptions import ConvergenceWarning
from sklearn.exceptions import DataConversionWarning
from sklearn.exceptions import SkipTestWarning
@@ -67,16 +66,6 @@ MULTI_OUTPUT = ['CCA', 'DecisionTreeRegressor', 'ElasticNet',
'RANSACRegressor', 'RadiusNeighborsRegressor',
'RandomForestRegressor', 'Ridge', 'RidgeCV']
-# Estimators with deprecated transform methods. Should be removed in 0.19 when
-# _LearntSelectorMixin is removed.
-DEPRECATED_TRANSFORM = [
- "RandomForestClassifier", "RandomForestRegressor", "ExtraTreesClassifier",
- "ExtraTreesRegressor", "DecisionTreeClassifier",
- "DecisionTreeRegressor", "ExtraTreeClassifier", "ExtraTreeRegressor",
- "LinearSVC", "SGDClassifier", "SGDRegressor", "Perceptron",
- "LogisticRegression", "LogisticRegressionCV",
- "GradientBoostingClassifier", "GradientBoostingRegressor"]
-
def _yield_non_meta_checks(name, Estimator):
yield check_estimators_dtypes
@@ -221,9 +210,8 @@ def _yield_all_checks(name, Estimator):
for check in _yield_regressor_checks(name, Estimator):
yield check
if issubclass(Estimator, TransformerMixin):
- if name not in DEPRECATED_TRANSFORM:
- for check in _yield_transformer_checks(name, Estimator):
- yield check
+ for check in _yield_transformer_checks(name, Estimator):
+ yield check
if issubclass(Estimator, ClusterMixin):
for check in _yield_clustering_checks(name, Estimator):
yield check
@@ -329,10 +317,6 @@ def set_testing_parameters(estimator):
# which is more feature than we have in most case.
estimator.set_params(k=1)
- if isinstance(estimator, NMF):
- if not isinstance(estimator, ProjectedGradientNMF):
- estimator.set_params(solver='cd')
-
class NotAnArray(object):
" An object that is convertable to an array"
@@ -424,8 +408,7 @@ def check_dtype_object(name, Estimator):
if hasattr(estimator, "predict"):
estimator.predict(X)
- if (hasattr(estimator, "transform") and
- name not in DEPRECATED_TRANSFORM):
+ if hasattr(estimator, "transform"):
estimator.transform(X)
try:
@@ -725,10 +708,7 @@ def check_pipeline_consistency(name, Estimator):
estimator.fit(X, y)
pipeline.fit(X, y)
- if name in DEPRECATED_TRANSFORM:
- funcs = ["score"]
- else:
- funcs = ["score", "fit_transform"]
+ funcs = ["score", "fit_transform"]
for func_name in funcs:
func = getattr(estimator, func_name, None)
@@ -751,11 +731,7 @@ def check_fit_score_takes_y(name, Estimator):
set_testing_parameters(estimator)
set_random_state(estimator)
- if name in DEPRECATED_TRANSFORM:
- funcs = ["fit", "score", "partial_fit", "fit_predict"]
- else:
- funcs = [
- "fit", "score", "partial_fit", "fit_predict", "fit_transform"]
+ funcs = ["fit", "score", "partial_fit", "fit_predict", "fit_transform"]
for func_name in funcs:
func = getattr(estimator, func_name, None)
if func is not None:
@@ -777,11 +753,7 @@ def check_estimators_dtypes(name, Estimator):
y = X_train_int_64[:, 0]
y = multioutput_estimator_convert_y_2d(name, y)
- if name in DEPRECATED_TRANSFORM:
- methods = ["predict", "decision_function", "predict_proba"]
- else:
- methods = [
- "predict", "transform", "decision_function", "predict_proba"]
+ methods = ["predict", "transform", "decision_function", "predict_proba"]
for X_train in [X_train_32, X_train_64, X_train_int_64, X_train_int_32]:
estimator = Estimator()
@@ -869,8 +841,7 @@ def check_estimators_nan_inf(name, Estimator):
raise AssertionError(error_string_predict, Estimator)
# transform
- if (hasattr(estimator, "transform") and
- name not in DEPRECATED_TRANSFORM):
+ if hasattr(estimator, "transform"):
try:
estimator.transform(X_train)
except ValueError as e:
@@ -888,11 +859,8 @@ def check_estimators_nan_inf(name, Estimator):
@ignore_warnings
def check_estimators_pickle(name, Estimator):
"""Test that we can pickle all estimators"""
- if name in DEPRECATED_TRANSFORM:
- check_methods = ["predict", "decision_function", "predict_proba"]
- else:
- check_methods = ["predict", "transform", "decision_function",
- "predict_proba"]
+ check_methods = ["predict", "transform", "decision_function",
+ "predict_proba"]
X, y = make_blobs(n_samples=30, centers=[[0, 0, 0], [1, 1, 1]],
random_state=0, n_features=2, cluster_std=0.1)
diff --git a/sklearn/utils/testing.py b/sklearn/utils/testing.py
index ce5465ad1f..9a62b3c6a9 100644
--- a/sklearn/utils/testing.py
+++ b/sklearn/utils/testing.py
@@ -15,8 +15,6 @@ import inspect
import pkgutil
import warnings
import sys
-import re
-import platform
import struct
import scipy as sp
@@ -636,28 +634,6 @@ def skip_if_32bit(func):
return run_test
-def if_not_mac_os(versions=('10.7', '10.8', '10.9'),
- message='Multi-process bug in Mac OS X >= 10.7 '
- '(see issue #636)'):
- """Test decorator that skips test if OS is Mac OS X and its
- major version is one of ``versions``.
- """
- warnings.warn("if_not_mac_os is deprecated in 0.17 and will be removed"
- " in 0.19: use the safer and more generic"
- " if_safe_multiprocessing_with_blas instead",
- DeprecationWarning)
- mac_version, _, _ = platform.mac_ver()
- skip = '.'.join(mac_version.split('.')[:2]) in versions
-
- def decorator(func):
- if skip:
- @wraps(func)
- def func(*args, **kwargs):
- raise SkipTest(message)
- return func
- return decorator
-
-
def if_safe_multiprocessing_with_blas(func):
"""Decorator for tests involving both BLAS calls and multiprocessing.
diff --git a/sklearn/utils/tests/test_class_weight.py b/sklearn/utils/tests/test_class_weight.py
index 1bf6377a48..a073eeafcf 100644
--- a/sklearn/utils/tests/test_class_weight.py
+++ b/sklearn/utils/tests/test_class_weight.py
@@ -12,17 +12,12 @@ from sklearn.utils.testing import assert_raises
from sklearn.utils.testing import assert_raise_message
from sklearn.utils.testing import assert_true
from sklearn.utils.testing import assert_equal
-from sklearn.utils.testing import assert_warns
def test_compute_class_weight():
# Test (and demo) compute_class_weight.
y = np.asarray([2, 2, 2, 3, 3, 4])
classes = np.unique(y)
- cw = assert_warns(DeprecationWarning,
- compute_class_weight, "auto", classes, y)
- assert_almost_equal(cw.sum(), classes.shape)
- assert_true(cw[0] < cw[1] < cw[2])
cw = compute_class_weight("balanced", classes, y)
# total effect of samples is preserved
@@ -35,11 +30,9 @@ def test_compute_class_weight_not_present():
# Raise error when y does not contain all class labels
classes = np.arange(4)
y = np.asarray([0, 0, 0, 1, 1, 2])
- assert_raises(ValueError, compute_class_weight, "auto", classes, y)
assert_raises(ValueError, compute_class_weight, "balanced", classes, y)
# Raise error when y has items not in classes
classes = np.arange(2)
- assert_raises(ValueError, compute_class_weight, "auto", classes, y)
assert_raises(ValueError, compute_class_weight, "balanced", classes, y)
assert_raises(ValueError, compute_class_weight, {0: 1., 1: 2.}, classes, y)
@@ -94,16 +87,11 @@ def test_compute_class_weight_invariance():
assert_array_almost_equal(logreg.coef_, logreg0.coef_)
-def test_compute_class_weight_auto_negative():
+def test_compute_class_weight_balanced_negative():
# Test compute_class_weight when labels are negative
# Test with balanced class labels.
classes = np.array([-2, -1, 0])
y = np.asarray([-1, -1, 0, 0, -2, -2])
- cw = assert_warns(DeprecationWarning, compute_class_weight, "auto",
- classes, y)
- assert_almost_equal(cw.sum(), classes.shape)
- assert_equal(len(cw), len(classes))
- assert_array_almost_equal(cw, np.array([1., 1., 1.]))
cw = compute_class_weight("balanced", classes, y)
assert_equal(len(cw), len(classes))
@@ -111,11 +99,6 @@ def test_compute_class_weight_auto_negative():
# Test with unbalanced class labels.
y = np.asarray([-1, 0, 0, -2, -2, -2])
- cw = assert_warns(DeprecationWarning, compute_class_weight, "auto",
- classes, y)
- assert_almost_equal(cw.sum(), classes.shape)
- assert_equal(len(cw), len(classes))
- assert_array_almost_equal(cw, np.array([0.545, 1.636, 0.818]), decimal=3)
cw = compute_class_weight("balanced", classes, y)
assert_equal(len(cw), len(classes))
@@ -124,15 +107,10 @@ def test_compute_class_weight_auto_negative():
assert_array_almost_equal(cw, [2. / 3, 2., 1.])
-def test_compute_class_weight_auto_unordered():
+def test_compute_class_weight_balanced_unordered():
# Test compute_class_weight when classes are unordered
classes = np.array([1, 0, 3])
y = np.asarray([1, 0, 0, 3, 3, 3])
- cw = assert_warns(DeprecationWarning, compute_class_weight, "auto",
- classes, y)
- assert_almost_equal(cw.sum(), classes.shape)
- assert_equal(len(cw), len(classes))
- assert_array_almost_equal(cw, np.array([1.636, 0.818, 0.545]), decimal=3)
cw = compute_class_weight("balanced", classes, y)
class_counts = np.bincount(y)[classes]
@@ -144,9 +122,6 @@ def test_compute_sample_weight():
# Test (and demo) compute_sample_weight.
# Test with balanced classes
y = np.asarray([1, 1, 1, 2, 2, 2])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
sample_weight = compute_sample_weight("balanced", y)
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
@@ -156,20 +131,14 @@ def test_compute_sample_weight():
# Test with column vector of balanced classes
y = np.asarray([[1], [1], [1], [2], [2], [2]])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
sample_weight = compute_sample_weight("balanced", y)
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
# Test with unbalanced classes
y = np.asarray([1, 1, 1, 2, 2, 2, 3])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- expected_auto = np.asarray([.6, .6, .6, .6, .6, .6, 1.8])
- assert_array_almost_equal(sample_weight, expected_auto)
sample_weight = compute_sample_weight("balanced", y)
- expected_balanced = np.array([0.7777, 0.7777, 0.7777, 0.7777, 0.7777, 0.7777, 2.3333])
+ expected_balanced = np.array([0.7777, 0.7777, 0.7777, 0.7777, 0.7777,
+ 0.7777, 2.3333])
assert_array_almost_equal(sample_weight, expected_balanced, decimal=4)
# Test with `None` weights
@@ -178,9 +147,6 @@ def test_compute_sample_weight():
# Test with multi-output of balanced classes
y = np.asarray([[1, 0], [1, 0], [1, 0], [2, 1], [2, 1], [2, 1]])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
sample_weight = compute_sample_weight("balanced", y)
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
@@ -191,9 +157,6 @@ def test_compute_sample_weight():
# Test with multi-output of unbalanced classes
y = np.asarray([[1, 0], [1, 0], [1, 0], [2, 1], [2, 1], [2, 1], [3, -1]])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, expected_auto ** 2)
sample_weight = compute_sample_weight("balanced", y)
assert_array_almost_equal(sample_weight, expected_balanced ** 2, decimal=3)
@@ -202,60 +165,38 @@ def test_compute_sample_weight_with_subsample():
# Test compute_sample_weight with subsamples specified.
# Test with balanced classes and all samples present
y = np.asarray([1, 1, 1, 2, 2, 2])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
sample_weight = compute_sample_weight("balanced", y, range(6))
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
# Test with column vector of balanced classes and all samples present
y = np.asarray([[1], [1], [1], [2], [2], [2]])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y)
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
sample_weight = compute_sample_weight("balanced", y, range(6))
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1.])
# Test with a subsample
y = np.asarray([1, 1, 1, 2, 2, 2])
- sample_weight = assert_warns(DeprecationWarning,
- compute_sample_weight, "auto", y, range(4))
- assert_array_almost_equal(sample_weight, [.5, .5, .5, 1.5, 1.5, 1.5])
sample_weight = compute_sample_weight("balanced", y, range(4))
assert_array_almost_equal(sample_weight, [2. / 3, 2. / 3,
2. / 3, 2., 2., 2.])
# Test with a bootstrap subsample
y = np.asarray([1, 1, 1, 2, 2, 2])
- sample_weight = assert_warns(DeprecationWarning, compute_sample_weight,
- "auto", y, [0, 1, 1, 2, 2, 3])
- expected_auto = np.asarray([1 / 3., 1 / 3., 1 / 3., 5 / 3., 5 / 3., 5 / 3.])
- assert_array_almost_equal(sample_weight, expected_auto)
sample_weight = compute_sample_weight("balanced", y, [0, 1, 1, 2, 2, 3])
expected_balanced = np.asarray([0.6, 0.6, 0.6, 3., 3., 3.])
assert_array_almost_equal(sample_weight, expected_balanced)
# Test with a bootstrap subsample for multi-output
y = np.asarray([[1, 0], [1, 0], [1, 0], [2, 1], [2, 1], [2, 1]])
- sample_weight = assert_warns(DeprecationWarning, compute_sample_weight,
- "auto", y, [0, 1, 1, 2, 2, 3])
- assert_array_almost_equal(sample_weight, expected_auto ** 2)
sample_weight = compute_sample_weight("balanced", y, [0, 1, 1, 2, 2, 3])
assert_array_almost_equal(sample_weight, expected_balanced ** 2)
# Test with a missing class
y = np.asarray([1, 1, 1, 2, 2, 2, 3])
- sample_weight = assert_warns(DeprecationWarning, compute_sample_weight,
- "auto", y, range(6))
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1., 0.])
sample_weight = compute_sample_weight("balanced", y, range(6))
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1., 0.])
# Test with a missing class for multi-output
y = np.asarray([[1, 0], [1, 0], [1, 0], [2, 1], [2, 1], [2, 1], [2, 2]])
- sample_weight = assert_warns(DeprecationWarning, compute_sample_weight,
- "auto", y, range(6))
- assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1., 0.])
sample_weight = compute_sample_weight("balanced", y, range(6))
assert_array_almost_equal(sample_weight, [1., 1., 1., 1., 1., 1., 0.])
@@ -270,7 +211,7 @@ def test_compute_sample_weight_errors():
assert_raises(ValueError, compute_sample_weight, "ni", y_)
assert_raises(ValueError, compute_sample_weight, "ni", y_, range(4))
- # Not "auto" for subsample
+ # Not "balanced" for subsample
assert_raises(ValueError,
compute_sample_weight, {1: 2, 2: 1}, y, range(4))
diff --git a/sklearn/utils/tests/test_validation.py b/sklearn/utils/tests/test_validation.py
index 9c57ab4e96..752af7e03f 100644
--- a/sklearn/utils/tests/test_validation.py
+++ b/sklearn/utils/tests/test_validation.py
@@ -122,9 +122,6 @@ def test_check_array():
X_csr = sp.csr_matrix(X)
assert_raises(TypeError, check_array, X_csr)
# ensure_2d
- assert_warns(DeprecationWarning, check_array, [0, 1, 2])
- X_array = check_array([0, 1, 2])
- assert_equal(X_array.ndim, 2)
X_array = check_array([0, 1, 2], ensure_2d=False)
assert_equal(X_array.ndim, 1)
# don't allow ndim > 3
@@ -338,12 +335,6 @@ def test_check_array_min_samples_and_features_messages():
msg = "Singleton array array(42) cannot be considered a valid collection."
assert_raise_message(TypeError, msg, check_array, 42, ensure_2d=False)
- # But this works if the input data is forced to look like a 2 array with
- # one sample and one feature:
- X_checked = assert_warns(DeprecationWarning, check_array, [42],
- ensure_2d=True)
- assert_array_equal(np.array([[42]]), X_checked)
-
# Simulate a model that would need at least 2 samples to be well defined
X = np.ones((1, 10))
y = np.ones(1)
diff --git a/sklearn/utils/validation.py b/sklearn/utils/validation.py
index deb98eef85..1a4b493e02 100644
--- a/sklearn/utils/validation.py
+++ b/sklearn/utils/validation.py
@@ -16,34 +16,16 @@ import scipy.sparse as sp
from ..externals import six
from ..utils.fixes import signature
-from .deprecation import deprecated
-from ..exceptions import DataConversionWarning as _DataConversionWarning
-from ..exceptions import NonBLASDotWarning as _NonBLASDotWarning
-from ..exceptions import NotFittedError as _NotFittedError
+from ..exceptions import NonBLASDotWarning
+from ..exceptions import NotFittedError
+from ..exceptions import DataConversionWarning
-@deprecated("DataConversionWarning has been moved into the sklearn.exceptions"
- " module. It will not be available here from version 0.19")
-class DataConversionWarning(_DataConversionWarning):
- pass
-
-
-@deprecated("NonBLASDotWarning has been moved into the sklearn.exceptions"
- " module. It will not be available here from version 0.19")
-class NonBLASDotWarning(_NonBLASDotWarning):
- pass
-
-
-@deprecated("NotFittedError has been moved into the sklearn.exceptions module."
- " It will not be available here from version 0.19")
-class NotFittedError(_NotFittedError):
- pass
-
FLOAT_DTYPES = (np.float64, np.float32, np.float16)
# Silenced by default to reduce verbosity. Turn on at runtime for
# performance profiling.
-warnings.simplefilter('ignore', _NonBLASDotWarning)
+warnings.simplefilter('ignore', NonBLASDotWarning)
def _assert_all_finite(X):
@@ -311,7 +293,7 @@ def check_array(array, accept_sparse=None, dtype="numeric", order=None,
Whether to raise an error on np.inf and np.nan in X.
ensure_2d : boolean (default=True)
- Whether to make X at least 2d.
+ Whether to raise a value error if X is not 2d.
allow_nd : boolean (default=False)
Whether to allow X.ndim > 2.
@@ -383,16 +365,10 @@ def check_array(array, accept_sparse=None, dtype="numeric", order=None,
if ensure_2d:
if array.ndim == 1:
- if ensure_min_samples >= 2:
- raise ValueError("%s expects at least 2 samples provided "
- "in a 2 dimensional array-like input"
- % estimator_name)
- warnings.warn(
- "Passing 1d arrays as data is deprecated in 0.17 and will "
- "raise ValueError in 0.19. Reshape your data either using "
+ raise ValueError(
+ "Got X with X.ndim=1. Reshape your data either using "
"X.reshape(-1, 1) if your data has a single feature or "
- "X.reshape(1, -1) if it contains a single sample.",
- DeprecationWarning)
+ "X.reshape(1, -1) if it contains a single sample.")
array = np.atleast_2d(array)
# To ensure that array flags are maintained
array = np.array(array, dtype=dtype, order=order, copy=copy)
@@ -426,7 +402,7 @@ def check_array(array, accept_sparse=None, dtype="numeric", order=None,
if warn_on_dtype and dtype_orig is not None and array.dtype != dtype_orig:
msg = ("Data with input dtype %s was converted to %s%s."
% (dtype_orig, array.dtype, context))
- warnings.warn(msg, _DataConversionWarning)
+ warnings.warn(msg, DataConversionWarning)
return array
@@ -556,7 +532,7 @@ def column_or_1d(y, warn=False):
warnings.warn("A column-vector y was passed when a 1d array was"
" expected. Please change the shape of y to "
"(n_samples, ), for example using ravel().",
- _DataConversionWarning, stacklevel=2)
+ DataConversionWarning, stacklevel=2)
return np.ravel(y)
raise ValueError("bad input shape {0}".format(shape))
@@ -686,8 +662,7 @@ def check_is_fitted(estimator, attributes, msg=None, all_or_any=all):
attributes = [attributes]
if not all_or_any([hasattr(estimator, attr) for attr in attributes]):
- # FIXME NotFittedError_ --> NotFittedError in 0.19
- raise _NotFittedError(msg % {'name': type(estimator).__name__})
+ raise NotFittedError(msg % {'name': type(estimator).__name__})
def check_non_negative(X, whom):
--
GitLab