Fix scikit-learn warnings.

examples/autoencoder/stacked_conv_autoencoders.py

@@ -3,7 +3,7 @@
 import theano
 import numpy as np
 from sklearn import datasets, metrics
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 from sklearn.preprocessing import OneHotEncoder
 from neupy import algorithms, layers, environment, surgery
 

examples/boltzmann_machine/rbm_faces_sampling.py

@@ -4,7 +4,7 @@
 import numpy as np
 import matplotlib.pyplot as plt
 from sklearn import datasets
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 from skimage.filters import threshold_adaptive
 from neupy import algorithms, environment
 from neupy.utils import asfloat

examples/cnn/mnist_cnn.py

@@ -1,7 +1,7 @@
 import theano
 import numpy as np
 from sklearn.preprocessing import OneHotEncoder
-from sklearn import cross_validation, metrics, datasets
+from sklearn import model_selection, metrics, datasets
 from neupy import algorithms, layers, environment
 
 
@@ -18,7 +18,7 @@
 n_samples = data.shape[0]
 data = data.reshape((n_samples, 1, 28, 28))
 
-x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+x_train, x_test, y_train, y_test = model_selection.train_test_split(
     data.astype(np.float32),
     target.astype(np.float32),
     train_size=(6 / 7.)

examples/mlp/boston_price_prediction.py

@@ -1,5 +1,5 @@
 from sklearn import datasets, preprocessing
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 import matplotlib.pyplot as plt
 from neupy import algorithms, layers, estimators, environment
 

examples/mlp/mnist_mlp.py

@@ -1,7 +1,7 @@
 import theano
 import numpy as np
 from sklearn.preprocessing import OneHotEncoder
-from sklearn import cross_validation, metrics, datasets
+from sklearn import model_selection, metrics, datasets
 from neupy import algorithms, layers, environment
 
 
@@ -17,7 +17,7 @@
 data = mnist.data / 255.
 data = data - data.mean(axis=0)
 
-x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+x_train, x_test, y_train, y_test = model_selection.train_test_split(
     data.astype(np.float32),
     target.astype(np.float32),
     train_size=(6 / 7.)

examples/mlp/rectangles_mlp.py

@@ -1,4 +1,4 @@
-from sklearn import cross_validation, metrics
+from sklearn import model_selection, metrics
 from skdata.larochelle_etal_2007 import dataset
 from neupy import algorithms, layers, environment
 
@@ -9,7 +9,7 @@
 rectangle_dataset.fetch(download_if_missing=True)
 
 data, target = rectangle_dataset.classification_task()
-x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+x_train, x_test, y_train, y_test = model_selection.train_test_split(
     data, target, train_size=0.5
 )
 

examples/rbfn/grnn_params_selection.py

@@ -2,7 +2,7 @@
 
 import numpy as np
 from sklearn import datasets, grid_search
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 from neupy import algorithms, estimators, environment
 
 

examples/rbfn/music_speech/train.py

@@ -12,7 +12,7 @@
 """
 import numpy as np
 from neupy import algorithms
-from sklearn import preprocessing, cross_validation, metrics, decomposition
+from sklearn import preprocessing, model_selection, metrics, decomposition
 import matplotlib.pyplot as plt
 from librosa.feature import mfcc
 from sklearn.utils import shuffle
@@ -67,10 +67,11 @@ def extract_features(data, n_fft=2048):
 
 print("\n> Train prediction")
 
-skf = cross_validation.StratifiedKFold(y_train, n_folds=5)
+skf = model_selection.StratifiedKFold(n_splits=5)
+skf_iterator = skf.split(x_train, y_train)
 scores = []
 
-for i, (train_index, test_index) in enumerate(skf, start=1):
+for i, (train_index, test_index) in enumerate(skf_iterator, start=1):
     print("\nK-fold #{}".format(i))
     pnnet = algorithms.PNN(std=std, verbose=False)
 

examples/rbfn/pnn_iris.py

@@ -1,6 +1,6 @@
 import numpy as np
 from sklearn import datasets
-from sklearn.cross_validation import StratifiedKFold
+from sklearn.model_selection import StratifiedKFold
 
 from neupy.algorithms import PNN
 
@@ -10,12 +10,12 @@
 target = dataset.target
 
 test_data_size = 10
-skfold = StratifiedKFold(target, test_data_size)
+skfold = StratifiedKFold(n_splits=test_data_size)
 avarage_result = 0
 
 print("> Start classify iris dataset")
 
-for i, (train, test) in enumerate(skfold, start=1):
+for i, (train, test) in enumerate(skfold.split(data, target), start=1):
     x_train, x_test = data[train], data[test]
     y_train, y_test = target[train], target[test]
 

neupy/algorithms/associative/hebb.py

@@ -23,7 +23,7 @@ class HebbRule(BaseStepAssociative):
     ----------
     decay_rate : float
         Decay rate is control your network weights. It helps network
-        'forgote' information and control weight sizes. Without this
+        'forget' information and control weight sizes. Without this
         parameter network weight will grow. Defaults to ``0.2``.
     {BaseAssociative.n_inputs}
     {BaseAssociative.n_outputs}

neupy/algorithms/ensemble/dan.py

@@ -35,7 +35,7 @@ class DynamicallyAveragedNetwork(BaseEnsemble):
     --------
     >>> import numpy as np
     >>> from sklearn import datasets, metrics
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms
     >>>
     >>> np.random.seed(50)

neupy/algorithms/ensemble/mixture_of_experts.py

@@ -37,7 +37,7 @@ class MixtureOfExperts(BaseEnsemble):
     --------
     >>> import numpy as np
     >>> from sklearn import datasets, preprocessing
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms, layers
     >>> from neupy.estimators import rmsle
     >>>

neupy/algorithms/gd/conjgrad.py

@@ -79,7 +79,7 @@ class ConjugateGradient(NoMultipleStepSelection, GradientDescent):
     Examples
     --------
     >>> from sklearn import datasets, preprocessing
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms, layers, estimators, environment
     >>>
     >>> environment.reproducible()

neupy/algorithms/gd/hessdiag.py

@@ -54,7 +54,7 @@ class HessianDiagonal(NoMultipleStepSelection, GradientDescent):
     Diabets dataset example
 
     >>> import numpy as np
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from sklearn import datasets, preprocessing
     >>> from neupy import algorithms, layers, environment
     >>> from neupy.estimators import rmsle

neupy/algorithms/rbfn/grnn.py

@@ -50,7 +50,7 @@ class GRNN(BaseNetwork, LazyLearningMixin):
     Examples
     --------
     >>> from sklearn import datasets, preprocessing
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms, estimators, environment
     >>>
     >>> environment.reproducible()

neupy/algorithms/rbfn/pnn.py

@@ -57,7 +57,7 @@ class PNN(BaseNetwork, LazyLearningMixin, MinibatchTrainingMixin):
     >>> import numpy as np
     >>>
     >>> from sklearn import datasets, metrics
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms, environment
     >>>
     >>> environment.reproducible()

neupy/algorithms/step_update/linear_search.py

@@ -31,7 +31,7 @@ class LinearSearch(SingleStepConfigurable):
     Examples
     --------
     >>> from sklearn import datasets, preprocessing
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>> from neupy import algorithms, layers, estimators, environment
     >>>
     >>> environment.reproducible()

requirements/test.txt

@@ -1,4 +1,4 @@
-scikit-learn>=0.15.2
+scikit-learn>=0.18.0
 pandas==0.16.0
 Pillow==3.0.0
 

site/2015/07/04/boston_house_prices_dataset.rst

@@ -243,7 +243,7 @@ We use 85% of data for train.
 
 .. code-block:: python
 
-    from sklearn.cross_validation import train_test_split
+    from sklearn.model_selection import train_test_split
     from neupy import environment
 
     environment.reproducible()

site/docs/quickstart.rst

@@ -15,7 +15,7 @@ First of all we need to load data.
 
 .. code-block:: python
 
-    >>> from sklearn import datasets, cross_validation
+    >>> from sklearn import datasets, model_selection
     >>> mnist = datasets.fetch_mldata('MNIST original')
     >>> data, target = mnist.data, mnist.target
 
@@ -43,7 +43,7 @@ samples for training and 10,000 for test.
 
     >>> from neupy import environment
     >>> import numpy as np
-    >>> from sklearn.cross_validation import train_test_split
+    >>> from sklearn.model_selection import train_test_split
     >>>
     >>> environment.reproducible()
     >>>

tests/algorithms/gd/test_levenberg_marquardt.py

@@ -2,7 +2,7 @@
 import theano
 import theano.tensor as T
 from sklearn import datasets, preprocessing
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 
 from neupy import algorithms, layers
 from neupy.utils import asfloat

tests/algorithms/gd/test_quickprop.py

@@ -1,6 +1,6 @@
 from functools import partial
 
-from sklearn import datasets, cross_validation, preprocessing
+from sklearn import datasets, model_selection, preprocessing
 
 from neupy import algorithms
 
@@ -16,7 +16,7 @@ def setUp(self):
                                                 random_state=33)
         target_scaler = preprocessing.MinMaxScaler()
         target = target_scaler.fit_transform(target.reshape(-1, 1))
-        self.data = cross_validation.train_test_split(data, target,
+        self.data = model_selection.train_test_split(data, target,
                                                       train_size=0.75)
         self.connection = (5, 10, 1)
 

tests/algorithms/rbfn/test_grnn.py

@@ -1,6 +1,6 @@
 import numpy as np
 from sklearn import datasets, metrics
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 
 from neupy import algorithms
 from base import BaseTestCase

tests/algorithms/rbfn/test_pnn.py

@@ -3,7 +3,7 @@
 import numpy as np
 from sklearn import datasets
 from sklearn import metrics
-from sklearn.cross_validation import StratifiedKFold, train_test_split
+from sklearn.model_selection import StratifiedKFold, train_test_split
 
 from neupy import algorithms
 from base import BaseTestCase
@@ -36,10 +36,10 @@ def test_simple_pnn(self):
         target = dataset.target
 
         test_data_size = 10
-        skfold = StratifiedKFold(target, test_data_size)
+        skfold = StratifiedKFold(n_splits=test_data_size)
         avarage_result = 0
 
-        for train, test in skfold:
+        for train, test in skfold.split(data, target):
             x_train, x_test = data[train], data[test]
             y_train, y_test = target[train], target[test]
 

tests/algorithms/steps/test_linear_search.py

@@ -1,6 +1,6 @@
 import numpy as np
 from sklearn import datasets, preprocessing
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 
 from neupy import algorithms, layers
 from neupy.estimators import rmsle

tests/compatibilities/test_pandas.py

@@ -1,6 +1,6 @@
 import pandas as pd
 from sklearn import datasets, preprocessing
-from sklearn.cross_validation import train_test_split
+from sklearn.model_selection import train_test_split
 from neupy import algorithms, layers, estimators
 
 from base import BaseTestCase

tests/compatibilities/test_sklearn_compatibility.py

@@ -1,8 +1,8 @@
 from operator import itemgetter
 
 import numpy as np
-from sklearn import datasets, preprocessing, metrics, grid_search
-from sklearn.cross_validation import train_test_split
+from sklearn import datasets, preprocessing, metrics, model_selection
+from sklearn.model_selection import train_test_split
 from sklearn.pipeline import Pipeline
 from neupy import algorithms, layers
 from neupy.estimators import rmsle
@@ -81,7 +81,7 @@ def scorer(network, X, y):
 
         self.assertAlmostEqual(0.513, error, places=3)
 
-        random_search = grid_search.RandomizedSearchCV(
+        random_search = model_selection.RandomizedSearchCV(
             grnnet,
             param_distributions={'std': np.arange(1e-2, 0.1, 1e-4)},
             n_iter=10,

tests/data.py

@@ -1,7 +1,7 @@
 import numpy as np
 
 from sklearn import datasets
-from sklearn.cross_validation import StratifiedShuffleSplit
+from sklearn.model_selection import StratifiedShuffleSplit
 
 
 xor_input_train = np.array([[-1, -1], [-1, 1], [1, -1], [1, 1]])
@@ -68,10 +68,10 @@ def simple_classification(n_samples=100, n_features=10, random_state=33):
     X, y = datasets.make_classification(n_samples=n_samples,
                                         n_features=n_features,
                                         random_state=random_state)
-    shuffle_split = StratifiedShuffleSplit(y, 1, train_size=0.6,
+    shuffle_split = StratifiedShuffleSplit(n_splits=1, train_size=0.6,
                                            random_state=random_state)
 
-    train_index, test_index = next(shuffle_split.__iter__())
+    train_index, test_index = next(shuffle_split.split(X, y))
     x_train, x_test = X[train_index], X[test_index]
     y_train, y_test = y[train_index], y[test_index]
 

tests/ensemble/test_dan.py

@@ -1,4 +1,4 @@
-from sklearn import datasets, cross_validation, metrics
+from sklearn import datasets, model_selection, metrics
 from neupy import algorithms, init
 from neupy.layers import Relu, Sigmoid, Input
 
@@ -9,7 +9,7 @@ class DANTestCase(BaseTestCase):
     def test_handle_errors(self):
         data, target = datasets.make_classification(300, n_features=4,
                                                     n_classes=2)
-        x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+        x_train, x_test, y_train, y_test = model_selection.train_test_split(
             data, target, train_size=0.7
         )
 
@@ -46,7 +46,7 @@ def test_handle_errors(self):
     def test_dan(self):
         data, target = datasets.make_classification(300, n_features=4,
                                                     n_classes=2)
-        x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+        x_train, x_test, y_train, y_test = model_selection.train_test_split(
             data, target, train_size=0.7
         )
 

tests/ensemble/test_mixtures_of_experts.py

@@ -1,5 +1,5 @@
 import numpy as np
-from sklearn import datasets, preprocessing, cross_validation
+from sklearn import datasets, preprocessing, model_selection
 from neupy import algorithms, layers
 from neupy.utils import asfloat
 from neupy.estimators import rmsle
@@ -121,7 +121,7 @@ def test_mixture_of_experts(self):
 
         input_scaler = preprocessing.MinMaxScaler((-1, 1))
         output_scaler = preprocessing.MinMaxScaler()
-        x_train, x_test, y_train, y_test = cross_validation.train_test_split(
+        x_train, x_test, y_train, y_test = model_selection.train_test_split(
             input_scaler.fit_transform(data),
             output_scaler.fit_transform(target.reshape(-1, 1)),
             train_size=0.8