Replace scikit-learn mixin classes with pyts mixin classes (johannfao…

…uzi#100)

pyts/approximation/dft.py

@@ -4,15 +4,16 @@
 # License: BSD-3-Clause
 
 import numpy as np
-from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.base import BaseEstimator
 from sklearn.feature_selection import f_classif
 from sklearn.utils.validation import check_array, check_is_fitted, check_X_y
 from math import ceil
 from warnings import warn
+from ..base import UnivariateTransformerMixin
 from ..preprocessing import StandardScaler
 
 
-class DiscreteFourierTransform(BaseEstimator, TransformerMixin):
+class DiscreteFourierTransform(BaseEstimator, UnivariateTransformerMixin):
     """Discrete Fourier Transform.
 
     Parameters

pyts/approximation/mcb.py

@@ -6,7 +6,8 @@
 import numpy as np
 from numba import njit, prange
 from scipy.stats import norm
-from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.base import BaseEstimator
+from ..base import UnivariateTransformerMixin
 from sklearn.tree import DecisionTreeClassifier
 from sklearn.utils.validation import check_array, check_is_fitted, check_X_y
 from sklearn.utils.multiclass import check_classification_targets
@@ -47,7 +48,7 @@ def _digitize(X, bins):
     return X_binned.astype('int64')
 
 
-class MultipleCoefficientBinning(BaseEstimator, TransformerMixin):
+class MultipleCoefficientBinning(BaseEstimator, UnivariateTransformerMixin):
     """Bin continuous data into intervals column-wise.
 
     Parameters

pyts/approximation/paa.py

@@ -6,8 +6,9 @@
 import numpy as np
 from math import ceil
 from numba import njit, prange
-from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.base import BaseEstimator
 from sklearn.utils.validation import check_array
+from ..base import UnivariateTransformerMixin
 from ..utils import segmentation
 
 
@@ -20,7 +21,8 @@ def _paa(X, n_samples, n_timestamps, start, end, n_timestamps_new):
     return X_paa
 
 
-class PiecewiseAggregateApproximation(BaseEstimator, TransformerMixin):
+class PiecewiseAggregateApproximation(BaseEstimator,
+                                      UnivariateTransformerMixin):
     """Piecewise Aggregate Approximation.
 
     Parameters

pyts/approximation/sax.py

@@ -4,12 +4,14 @@
 # License: BSD-3-Clause
 
 import numpy as np
-from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.base import BaseEstimator
 from sklearn.utils.validation import check_array
+from ..base import UnivariateTransformerMixin
 from ..preprocessing import KBinsDiscretizer
 
 
-class SymbolicAggregateApproximation(BaseEstimator, TransformerMixin):
+class SymbolicAggregateApproximation(BaseEstimator,
+                                     UnivariateTransformerMixin):
     """Symbolic Aggregate approXimation.
 
     Parameters

pyts/approximation/sfa.py

@@ -3,14 +3,15 @@
 # Author: Johann Faouzi <[email protected]>
 # License: BSD-3-Clause
 
-from sklearn.base import BaseEstimator, TransformerMixin
+from sklearn.base import BaseEstimator
 from sklearn.pipeline import Pipeline
 from sklearn.utils.validation import check_is_fitted
 from .dft import DiscreteFourierTransform
 from .mcb import MultipleCoefficientBinning
+from ..base import UnivariateTransformerMixin
 
 
-class SymbolicFourierApproximation(BaseEstimator, TransformerMixin):
+class SymbolicFourierApproximation(BaseEstimator, UnivariateTransformerMixin):
     """Symbolic Fourier Approximation.
 
     Parameters

pyts/bag_of_words/bow.py

@@ -11,11 +11,12 @@
 import warnings
 from ..approximation import (
     PiecewiseAggregateApproximation, SymbolicAggregateApproximation)
+from ..base import UnivariateTransformerMixin
 from ..preprocessing import StandardScaler
 from ..utils.utils import _windowed_view
 
 
-class WordExtractor(BaseEstimator, TransformerMixin):
+class WordExtractor(BaseEstimator, UnivariateTransformerMixin):
     r"""Transform discretized time series into sequences of words.
 
     Parameters

pyts/base.py

@@ -0,0 +1,128 @@
+"""Base classes for all estimators."""
+
+# Author: Johann Faouzi <[email protected]>
+# License: BSD-3-Clause
+
+from sklearn.metrics import accuracy_score
+
+
+class UnivariateTransformerMixin:
+    """Mixin class for all univariate transformers in pyts."""
+
+    def fit_transform(self, X, y=None, **fit_params):
+        """Fit to data, then transform it.
+
+        Fits transformer to `X` and `y` with optional parameters `fit_params`
+        and returns a transformed version of `X`.
+
+        Parameters
+        ----------
+        X : array-like, shape = (n_samples, n_timestamps)
+            Univariate time series.
+
+        y : None or array-like, shape = (n_samples,) (default = None)
+            Target values (None for unsupervised transformations).
+
+        **fit_params : dict
+            Additional fit parameters.
+
+        Returns
+        -------
+        X_new : array
+            Transformed array.
+
+        """  # noqa: E501
+        if y is None:
+            # fit method of arity 1 (unsupervised transformation)
+            return self.fit(X, **fit_params).transform(X)
+        else:
+            # fit method of arity 2 (supervised transformation)
+            return self.fit(X, y, **fit_params).transform(X)
+
+
+class MultivariateTransformerMixin:
+    """Mixin class for all multivariate transformers in pyts."""
+
+    def fit_transform(self, X, y=None, **fit_params):
+        """Fit to data, then transform it.
+
+        Fits transformer to `X` and `y` with optional parameters `fit_params`
+        and returns a transformed version of `X`.
+
+        Parameters
+        ----------
+        X : array-like, shape = (n_samples, n_features, n_timestamps)
+            Multivariate time series.
+
+        y : None or array-like, shape = (n_samples,) (default = None)
+            Target values (None for unsupervised transformations).
+
+        **fit_params : dict
+            Additional fit parameters.
+
+        Returns
+        -------
+        X_new : array
+            Transformed array.
+
+        """  # noqa: E501
+        if y is None:
+            # fit method of arity 1 (unsupervised transformation)
+            return self.fit(X, **fit_params).transform(X)
+        else:
+            # fit method of arity 2 (supervised transformation)
+            return self.fit(X, y, **fit_params).transform(X)
+
+
+class UnivariateClassifierMixin:
+    """Mixin class for all univariate classifiers in pyts."""
+
+    def score(self, X, y, sample_weight=None):
+        """
+        Return the mean accuracy on the given test data and labels.
+
+        Parameters
+        ----------
+        X : array-like, shape = (n_samples, n_timestamps)
+            Univariate time series.
+
+        y : array-like, shape = (n_samples,)
+            True labels for `X`.
+
+        sample_weight : None or array-like, shape = (n_samples,) (default = None)
+            Sample weights.
+
+        Returns
+        -------
+        score : float
+            Mean accuracy of ``self.predict(X)`` with regards to `y`.
+
+        """  # noqa: E501
+        return accuracy_score(y, self.predict(X), sample_weight=sample_weight)
+
+
+class MultivariateClassifierMixin:
+    """Mixin class for all multivariate classifiers in pyts."""
+
+    def score(self, X, y, sample_weight=None):
+        """
+        Return the mean accuracy on the given test data and labels.
+
+        Parameters
+        ----------
+        X : array-like, shape = (n_samples, n_features, n_timestamps)
+            Multivariate time series.
+
+        y : array-like, shape = (n_samples,)
+            True labels for `X`.
+
+        sample_weight : None or array-like, shape = (n_samples,) (default = None)
+            Sample weights.
+
+        Returns
+        -------
+        score : float
+            Mean accuracy of ``self.predict(X)`` with regards to `y`.
+
+        """  # noqa: E501
+        return accuracy_score(y, self.predict(X), sample_weight=sample_weight)

pyts/classification/bossvs.py

@@ -7,15 +7,16 @@
 from math import ceil
 from sklearn.utils.validation import check_array, check_X_y, check_is_fitted
 from sklearn.utils.multiclass import check_classification_targets
-from sklearn.base import BaseEstimator, ClassifierMixin
+from sklearn.base import BaseEstimator
 from sklearn.metrics.pairwise import cosine_similarity
 from sklearn.preprocessing import LabelEncoder
 from sklearn.feature_extraction.text import TfidfVectorizer
 from ..approximation import SymbolicFourierApproximation
+from ..base import UnivariateClassifierMixin
 from ..utils.utils import _windowed_view
 
 
-class BOSSVS(BaseEstimator, ClassifierMixin):
+class BOSSVS(BaseEstimator, UnivariateClassifierMixin):
     """Bag-of-SFA Symbols in Vector Space.
 
     Each time series is transformed into an histogram using the

pyts/classification/knn.py

@@ -3,16 +3,17 @@
 # Author: Johann Faouzi <[email protected]>
 # License: BSD-3-Clause
 
-from sklearn.base import BaseEstimator, ClassifierMixin
+from sklearn.base import BaseEstimator
 from sklearn.neighbors import KNeighborsClassifier as SklearnKNN
 from sklearn.preprocessing import LabelEncoder
 from sklearn.utils.validation import check_X_y, check_is_fitted
+from ..base import UnivariateClassifierMixin
 from ..metrics import boss, dtw, sakoe_chiba_band, itakura_parallelogram
 from ..metrics.dtw import (_dtw_classic, _dtw_region, _dtw_fast,
                            _dtw_multiscale)
 
 
-class KNeighborsClassifier(BaseEstimator, ClassifierMixin):
+class KNeighborsClassifier(BaseEstimator, UnivariateClassifierMixin):
     """k-nearest neighbors classifier.
 
     Parameters

pyts/classification/learning_shapelets.py

@@ -5,11 +5,9 @@
 
 from itertools import chain
 from math import ceil
-
 from numba import njit, prange
 import numpy as np
-
-from sklearn.base import BaseEstimator, ClassifierMixin
+from sklearn.base import BaseEstimator
 from sklearn.cluster import KMeans
 from sklearn.exceptions import ConvergenceWarning
 from sklearn.multiclass import OneVsOneClassifier, OneVsRestClassifier
@@ -19,10 +17,9 @@
 from sklearn.utils.validation import (
     check_is_fitted, check_random_state, check_X_y, _check_sample_weight)
 from sklearn.utils.multiclass import check_classification_targets
-
 import warnings
-
 from ..utils.utils import _windowed_view
+from ..base import UnivariateClassifierMixin
 
 
 @njit(fastmath=True)
@@ -305,7 +302,7 @@ def _grad_shapelets(X, y, n_classes, weights, shapelets, lengths, alpha,
     return gradients
 
 
-class CrossEntropyLearningShapelets(BaseEstimator, ClassifierMixin):
+class CrossEntropyLearningShapelets(BaseEstimator, UnivariateClassifierMixin):
     """Learning Shapelets algorithm with cross-entropy loss.
 
     Parameters
@@ -811,7 +808,7 @@ def _check_params(self, X, y, y_ind, classes, sample_weight):
         return n_shapelets_per_size, min_shapelet_length, sample_weight, rng
 
 
-class LearningShapelets(BaseEstimator, ClassifierMixin):
+class LearningShapelets(BaseEstimator, UnivariateClassifierMixin):
     """Learning Shapelets algorithm.
 
     This estimator consists of two steps: computing the distances between the

pyts/classification/saxvsm.py

@@ -5,14 +5,15 @@
 
 from sklearn.utils.validation import check_X_y, check_array, check_is_fitted
 from sklearn.utils.multiclass import check_classification_targets
-from sklearn.base import BaseEstimator, ClassifierMixin
+from sklearn.base import BaseEstimator
 from sklearn.metrics.pairwise import cosine_similarity
 from sklearn.preprocessing import LabelEncoder
 from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
 from ..bag_of_words import BagOfWords
+from ..base import UnivariateClassifierMixin
 
 
-class SAXVSM(BaseEstimator, ClassifierMixin):
+class SAXVSM(BaseEstimator, UnivariateClassifierMixin):
     """Classifier based on SAX-VSM representation and tf-idf statistics.
 
     Time series are first transformed into bag of words using Symbolic

pyts/classification/time_series_forest.py

@@ -6,11 +6,12 @@
 from math import ceil
 from numba import njit
 import numpy as np
-from sklearn.base import BaseEstimator, ClassifierMixin, TransformerMixin
+from sklearn.base import BaseEstimator
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.pipeline import Pipeline
 from sklearn.utils.validation import (
     check_array, check_is_fitted, check_random_state)
+from ..base import UnivariateClassifierMixin, UnivariateTransformerMixin
 
 
 @njit()
@@ -35,7 +36,7 @@ def extract_features(X, n_samples, n_windows, indices):
     return X_new
 
 
-class WindowFeatureExtractor(BaseEstimator, TransformerMixin):
+class WindowFeatureExtractor(BaseEstimator, UnivariateTransformerMixin):
     """Feature extractor over a window.
 
     This transformer extracts 3 features from each window: the mean, the
@@ -174,7 +175,7 @@ def _check_params(self, X):
         return n_windows, min_window_size, rng
 
 
-class TimeSeriesForest(BaseEstimator, ClassifierMixin):
+class TimeSeriesForest(BaseEstimator, UnivariateClassifierMixin):
     """A random forest classifier for time series.
 
     A random forest is a meta estimator that fits a number of decision tree
@@ -376,7 +377,7 @@ class TimeSeriesForest(BaseEstimator, ClassifierMixin):
            Forest for Classification and Feature Extraction".
            Information Sciences, 239, 142-153 (2013).
 
-    .. [2] L. Breiman, "Random Forests", Machine Learning, 45(1), 5-32, 2001.
+    .. [2] Leo Breiman, "Random Forests", Machine Learning, 45(1), 5-32, 2001.
 
     """  # noqa: E501
     def __init__(self,