add TimeSeriesScaleMeanMaxVariance

Author: tonylee2016

tslearn/preprocessing/__init__.py

@@ -6,11 +6,13 @@
 from .preprocessing import (
     TimeSeriesScalerMeanVariance,
     TimeSeriesScalerMinMax,
-    TimeSeriesResampler
+    TimeSeriesResampler,
+    TimeSeriesScaleMeanMaxVariance
 )
 
 __all__ = [
     "TimeSeriesResampler",
     "TimeSeriesScalerMinMax",
-    "TimeSeriesScalerMeanVariance"
+    "TimeSeriesScalerMeanVariance",
+    "TimeSeriesScaleMeanMaxVariance",
 ]

tslearn/preprocessing/preprocessing.py

@@ -296,3 +296,91 @@ def transform(self, X, y=None, **kwargs):
 
     def _more_tags(self):
         return {'allow_nan': True}
+
+
+class TimeSeriesScaleMeanMaxVariance(TransformerMixin, TimeSeriesBaseEstimator):
+    """Scaler for time series. Scales time series so that their mean (resp.
+    standard deviation) in the signal with the max amplitue  is
+    mu (resp. std). The scaling relationships between each signal are preserved
+    This is supplement to the TimeSeriesScalerMeanVariance method
+
+    Parameters
+    ----------
+    mu : float (default: 0.)
+        Mean of the output time series.
+    std : float (default: 1.)
+        Standard deviation of the output time series.
+
+    Notes
+    -----
+        This method requires a dataset of equal-sized time series.
+
+        NaNs within a time series are ignored when calculating mu and std.
+    """
+
+    def __init__(self, mu=0., std=1.):
+        self.mu = mu
+        self.std = std
+
+    def fit(self, X, y=None, **kwargs):
+        """A dummy method such that it complies to the sklearn requirements.
+        Since this method is completely stateless, it just returns itself.
+
+        Parameters
+        ----------
+        X
+            Ignored
+
+        Returns
+        -------
+        self
+        """
+        X = check_array(X, allow_nd=True, force_all_finite=False)
+        X = to_time_series_dataset(X)
+        self._X_fit_dims = X.shape
+        return self
+
+    def fit_transform(self, X, y=None, **kwargs):
+        """Fit to data, then transform it.
+
+        Parameters
+        ----------
+        X : array-like of shape (n_ts, sz, d)
+            Time series dataset to be rescaled.
+
+        Returns
+        -------
+        numpy.ndarray
+            Resampled time series dataset.
+        """
+        return self.fit(X).transform(X)
+
+    def transform(self, X, y=None, **kwargs):
+        """Fit to data, then transform it.
+
+        Parameters
+        ----------
+        X : array-like of shape (n_ts, sz, d)
+            Time series dataset to be rescaled
+
+        Returns
+        -------
+        numpy.ndarray
+            Rescaled time series dataset
+        """
+        check_is_fitted(self, '_X_fit_dims')
+        X = check_array(X, allow_nd=True, force_all_finite=False)
+        X_ = to_time_series_dataset(X)
+        X_ = check_dims(X_, X_fit_dims=self._X_fit_dims, extend=False)
+        mean_t = numpy.nanmean(X_, axis=1, keepdims=True)
+        std_t = numpy.nanstd(X_, axis=1, keepdims=True)
+        # retain the scaling relation cross the signals,
+        # the max std_t is set to self.std
+        max_std = max(std_t)
+        if max_std ==0.: max_std = 1
+        X_ = (X_ - mean_t) * self.std / max_std + self.mu
+
+        return X_
+
+    def _more_tags(self):
+        return {'allow_nan': True}