8000 joblib 0.12.2 · scikit-learn/scikit-learn@4fc099d · GitHub
[go: up one dir, main page]
Skip to content

Commit 4fc099d

Browse files
ogriselogrisel
committed
joblib 0.12.2
1 parent eb453b6 commit 4fc099d

File tree

107 files changed

+9162
-1695
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

107 files changed

+9162
-1695
lines changed

‎build_tools/travis/install.sh

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -100,7 +100,8 @@ elif [[ "$DISTRIB" == "scipy-dev" ]]; then
100100
fi
101101

102102
if [[ "$COVERAGE" == "true" ]]; then
103-
pip install coverage codecov
103+
pip install hg+http://bitbucket.org/ogrisel/coverage.py/@fix-thread-safety#egg=coverage
104+
pip install codecov
104105
fi
105106

106107
if [[ "$TEST_DOCSTRINGS" == "true" ]]; then

‎doc/modules/compose.rst

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -342,7 +342,7 @@ and ``value`` is an estimator object::
342342
>>> estimators = [('linear_pca', PCA()), ('kernel_pca', KernelPCA())]
343343
>>> combined = FeatureUnion(estimators)
344344
>>> combined # doctest: +NORMALIZE_WHITESPACE, +ELLIPSIS
345-
FeatureUnion(n_jobs=1,
345+
FeatureUnion(n_jobs=None,
346346
transformer_list=[('linear_pca', PCA(copy=True,...)),
347347
('kernel_pca', KernelPCA(alpha=1.0,...))],
348348
transformer_weights=None)
@@ -357,7 +357,7 @@ and ignored by setting to ``None``::
357357

358358
>>> combined.set_params(kernel_pca=None)
359359
... # doctest: +NORMALIZE_WHITESPACE, +ELLIPSIS
360-
FeatureUnion(n_jobs=1,
360+
FeatureUnion(n_jobs=None,
361361
transformer_list=[('linear_pca', PCA(copy=True,...)),
362362
('kernel_pca', None)],
363363
transformer_weights=None)
@@ -423,7 +423,7 @@ By default, the remaining rating columns are ignored (``remainder='drop'``)::
423423
... remainder='drop')
424424

425425
>>> column_trans.fit(X) # doctest: +NORMALIZE_WHITESPACE +ELLIPSIS
426-
ColumnTransformer(n_jobs=1, remainder='drop', sparse_threshold=0.3,
426+
ColumnTransformer(n_jobs=None, remainder='drop', sparse_threshold=0.3,
427427
transformer_weights=None,
428428
transformers=...)
429429

@@ -496,7 +496,7 @@ above example would be::
496496
... ('city', CountVectorizer(analyzer=lambda x: [x])),
497497
... ('title', CountVectorizer()))
498498
>>> column_trans # doctest: +NORMALIZE_WHITESPACE +ELLIPSIS
499-
ColumnTransformer(n_jobs=1, remainder='drop', sparse_threshold=0.3,
499+
ColumnTransformer(n_jobs=None, remainder='drop', sparse_threshold=0.3,
500500
transformer_weights=None,
501501
transformers=[('countvectorizer-1', ...)
502502

‎doc/modules/kernel_approximation.rst

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -64,10 +64,9 @@ a linear algorithm, for example a linear SVM::
6464
SGDClassifier(alpha=0.0001, average=False, class_weight=None,
6565
early_stopping=False, epsilon=0.1, eta0=0.0, fit_intercept=True,
6666
l1_ratio=0.15, learning_rate='optimal', loss='hinge', max_iter=5,
67-
n_iter=None, n_iter_no_change=5, n_jobs=1, penalty='l2',
67+
n_iter=None, n_iter_no_change=5, n_jobs=None, penalty='l2',
6868
power_t=0.5, random_state=None, shuffle=True, tol=None,
6969
validation_fraction=0.1, verbose=0, warm_start=False)
70-
7170
>>> clf.score(X_features, y)
7271
1.0
7372

‎doc/modules/linear_model.rst

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -45,7 +45,9 @@ and will store the coefficients :math:`w` of the linear model in its
4545
>>> from sklearn import linear_model
4646
>>> reg = linear_model.LinearRegression()
4747
>>> reg.fit ([[0, 0], [1, 1], [2, 2]], [0, 1, 2])
48-
LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)
48+
... # doctest: +NORMALIZE_WHITESPACE
49+
LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,
50+
normalize=False)
4951
>>> reg.coef_
5052
array([0.5, 0.5])
5153

‎doc/modules/sgd.rst

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -64,7 +64,7 @@ for the training samples::
6464
SGDClassifier(alpha=0.0001, average=False, class_weight=None,
6565
early_stopping=False, epsilon=0.1, eta0=0.0, fit_intercept=True,
6666
l1_ratio=0.15, learning_rate='optimal', loss='hinge', max_iter=5,
67-
n_iter=None, n_iter_no_change=5, n_jobs=1, penalty='l2',
67+
n_iter=None, n_iter_no_change=5, n_jobs=None, penalty='l2',
6868
power_t=0.5, random_state=None, shuffle=True, tol=None,
6969
validation_fraction=0.1, verbose=0, warm_start=False)
7070

‎doc/tutorial/statistical_inference/model_selection.rst

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -269,9 +269,9 @@ parameter automatically by cross-validation::
269269
>>> y_diabetes = diabetes.target
270270
>>> lasso.fit(X_diabetes, y_diabetes)
271271
LassoCV(alphas=None, copy_X=True, cv=3, eps=0.001, fit_intercept=True,
272-
max_iter=1000, n_alphas=100, n_jobs=1, normalize=False, positive=False,
273-
precompute='auto', random_state=None, selection='cyclic', tol=0.0001,
274-
verbose=False)
272+
max_iter=1000, n_alphas=100, n_jobs=None, normalize=False,
273+
positive=False, precompute='auto', random_state=None,
274+
selection='cyclic', tol=0.0001, verbose=False)
275275
>>> # The estimator chose automatically its lambda:
276276
>>> lasso.alpha_ # doctest: +ELLIPSIS
277277
0.01229...

‎doc/tutorial/statistical_inference/supervised_learning.rst

Lines changed: 12 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -95,7 +95,7 @@ Scikit-learn documentation for more information about this type of classifier.)
9595
>>> knn = KNeighborsClassifier()
9696
>>> knn.fit(iris_X_train, iris_y_train) # doctest: +NORMALIZE_WHITESPACE
9797
KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',
98-
metric_params=None, n_jobs=1, n_neighbors=5, p=2,
98+
metric_params=None, n_jobs=None, n_neighbors=5, p=2,
9999
weights='uniform')
100100
>>> knn.predict(iris_X_test)
101101
array([1, 2, 1, 0, 0, 0, 2, 1, 2, 0])
@@ -176,13 +176,16 @@ Linear models: :math:`y = X\beta + \epsilon`
176176
>>> from sklearn import linear_model
177177
>>> regr = linear_model.LinearRegression()
178178
>>> regr.fit(diabetes_X_train, diabetes_y_train)
179-
LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)
179+
... # doctest: +NORMALIZE_WHITESPACE
180+
LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,
181+
normalize=False)
180182
>>> print(regr.coef_)
181183
[ 0.30349955 -237.63931533 510.53060544 327.73698041 -814.13170937
182184
492.81458798 102.84845219 184.60648906 743.51961675 76.09517222]
183185

184186
>>> # The mean square error
185-
>>> np.mean((regr.predict(diabetes_X_test)-diabetes_y_test)**2)# doctest: +ELLIPSIS
187+
>>> np.mean((regr.predict(diabetes_X_test)-diabetes_y_test)**2)
188+
... # doctest: +ELLIPSIS
186189
2004.56760268...
187190

188191
>>> # Explained variance score: 1 is perfect prediction
@@ -257,8 +260,11 @@ diabetes dataset rather than our synthetic data::
257260
>>> from __future__ import print_function
258261
>>> print([regr.set_params(alpha=alpha
259262
... ).fit(diabetes_X_train, diabetes_y_train,
260-
... ).score(diabetes_X_test, diabetes_y_test) for alpha in alphas]) # doctest: +ELLIPSIS
261-
[0.5851110683883..., 0.5852073015444..., 0.5854677540698..., 0.5855512036503..., 0.5830717085554..., 0.57058999437...]
263+
... ).score(diabetes_X_test, diabetes_y_test)
264+
... for alpha in alphas])
265+
... # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE
266+
[0.5851110683883..., 0.5852073015444..., 0.5854677540698...,
267+
0.5855512036503..., 0.5830717085554..., 0.57058999437...]
262268

263269

264270
.. note::
@@ -372,7 +378,7 @@ function or **logistic** function:
372378
>>> logistic.fit(iris_X_train, iris_y_train)
373379
LogisticRegression(C=100000.0, class_weight=None, dual=False,
374380
fit_intercept=True, intercept_scaling=1, max_iter=100,
375-
multi_class='ovr', n_jobs=1, penalty='l2', random_state=None,
381+
multi_class='ovr', n_jobs=None, penalty='l2', random_state=None,
376382
solver='liblinear', tol=0.0001, verbose=0, warm_start=False)
377383

378384
This is known as :class:`LogisticRegression`.

‎sklearn/cluster/bicluster.py

Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -93,7 +93,7 @@ class BaseSpectral(six.with_metaclass(ABCMeta, BaseEstimator,
9393
@abstractmethod
9494
def __init__(self, n_clusters=3, svd_method="randomized",
9595
n_svd_vecs=None, mini_batch=False, init="k-means++",
96-
n_init=10, n_jobs=1, random_state=None):
96+
n_init=10, n_jobs=None, random_state=None):
9797
self.n_clusters = n_clusters
9898
self.svd_method = svd_method
9999
self.n_svd_vecs = n_svd_vecs
@@ -271,7 +271,7 @@ class SpectralCoclustering(BaseSpectral):
271271
array([0, 0], dtype=int32)
272272
>>> clustering # doctest: +NORMALIZE_WHITESPACE
273273
SpectralCoclustering(init='k-means++', mini_batch=False, n_clusters=2,
274-
n_init=10, n_jobs=1, n_svd_vecs=None, random_state=0,
274+
n_init=10, n_jobs=None, n_svd_vecs=None, random_state=0,
275275
svd_method='randomized')
276276
277277
References
@@ -284,7 +284,7 @@ class SpectralCoclustering(BaseSpectral):
284284
"""
285285
def __init__(self, n_clusters=3, svd_method='randomized',
286286
n_svd_vecs=None, mini_batch=False, init='k-means++',
287-
n_init=10, n_jobs=1, random_state=None):
287+
n_init=10, n_jobs=None, random_state=None):
288288
super(SpectralCoclustering, self).__init__(n_clusters,
289289
svd_method,
290290
n_svd_vecs,
@@ -419,7 +419,7 @@ class SpectralBiclustering(BaseSpectral):
419419
>>> clustering # doctest: +NORMALIZE_WHITESPACE
420420
SpectralBiclustering(init='k-means++', method='bistochastic',
421421
mini_batch=False, n_best=3, n_clusters=2, n_components=6,
422-
n_init=10, n_jobs=1, n_svd_vecs=None, random_state=0,
422+
n_init=10, n_jobs=None, n_svd_vecs=None, random_state=0,
423423
svd_method='randomized')
424424
425425
References
@@ -433,7 +433,7 @@ class SpectralBiclustering(BaseSpectral):
433433
def __init__(self, n_clusters=3, method='bistochastic',
434434
n_components=6, n_best=3, svd_method='randomized',
435435
n_svd_vecs=None, mini_batch=False, init='k-means++',
436-
n_init=10, n_jobs=1, random_state=None):
436+
n_init=10, n_jobs=None, random_state=None):
437437
super(SpectralBiclustering, self).__init__(n_clusters,
438438
svd_method,
439439
n_svd_vecs,

‎sklearn/cluster/dbscan_.py

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -20,7 +20,8 @@
2020

2121

2222
def dbscan(X, eps=0.5, min_samples=5, metric='minkowski', metric_params=None,
23-
algorithm='auto', leaf_size=30, p=2, sample_weight=None, n_jobs=1):
23+
algorithm='auto', leaf_size=30, p=2, sample_weight=None,
24+
n_jobs=None):
2425
"""Perform DBSCAN clustering from vector array or distance matrix.
2526
2627
Read more in the :ref:`User Guide <dbscan>`.
@@ -255,7 +256,7 @@ class DBSCAN(BaseEstimator, ClusterMixin):
255256
array([ 0, 0, 0, 1, 1, -1])
256257
>>> clustering # doctest: +NORMALIZE_WHITESPACE
257258
DBSCAN(algorithm='auto', eps=3, leaf_size=30, metric='euclidean',
258-
metric_params=None, min_samples=2, n_jobs=1, p=None)
259+
metric_params=None, min_samples=2, n_jobs=None, p=None)
259260
260261
See also
261262
--------
@@ -296,7 +297,7 @@ class DBSCAN(BaseEstimator, ClusterMixin):
296297

297298
def __init__(self, eps=0.5, min_samples=5, metric='euclidean',
298299
metric_params=None, algorithm='auto', leaf_size=30, p=None,
299-
n_jobs=1):
300+
n_jobs=None):
300301
self.eps = eps
301302
self.min_samples = min_samples
302303
self.metric = metric

‎sklearn/cluster/k_means_.py

Lines changed: 6 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -25,12 +25,13 @@
2525
from ..utils.sparsefuncs import mean_variance_axis
2626
from ..utils.validation import _num_samples
2727
from ..utils import check_array
28-
from ..utils import check_random_state
2928
from ..utils import gen_batches
29+
from ..utils import check_random_state
3030
from ..utils.validation import check_is_fitted
3131
from ..utils.validation import FLOAT_DTYPES
3232
from ..utils import Parallel
3333
from ..utils import delayed
34+
from ..utils import effective_n_jobs
3435
from ..externals.six import string_types
3536
from ..exceptions import ConvergenceWarning
3637
from . import _k_means
@@ -184,8 +185,8 @@ def _check_sample_weight(X, sample_weight):
184185

185186
def k_means(X, n_clusters, sample_weight=None, init='k-means++',
186187
precompute_distances='auto', n_init=10, max_iter=300,
187-
verbose=False, tol=1e-4, random_state=None, copy_x=True, n_jobs=1,
188-
algorithm="auto", return_n_iter=False):
188+
verbose=False, tol=1e-4, random_state=None, copy_x=True,
189+
n_jobs=None, algorithm="auto", return_n_iter=False):
189190
"""K-means clustering algorithm.
190191
191192
Read more in the :ref:`User Guide <k_means>`.
@@ -368,7 +369,7 @@ def k_means(X, n_clusters, sample_weight=None, init='k-means++',
368369
else:
369370
raise ValueError("Algorithm must be 'auto', 'full' or 'elkan', got"
370371
" %s" % str(algorithm))
371-
if n_jobs == 1:
372+
if effective_n_jobs(n_jobs):
372373
# For a single thread, less memory is needed if we just store one set
373374
# of the best results (as opposed to one set per run per thread).
374375
for it in range(n_init):
@@ -913,7 +914,7 @@ class KMeans(BaseEstimator, ClusterMixin, TransformerMixin):
913914
def __init__(self, n_clusters=8, init='k-means++', n_init=10,
914915
max_iter=300, tol=1e-4, precompute_distances='auto',
915916
verbose=0, random_state=None, copy_x=True,
916-
n_jobs=1, algorithm='auto'):
917+
n_jobs=None, algorithm='auto'):
917918

918919
self.n_clusters = n_clusters
919920
self.init = init

0 commit comments

Comments
 (0)
0