c68ec2e70731f601f630eb1816c51d7ee4ef0853,tslearn/clustering.py,KShape,fit,#KShape#Any#Any#Any#,737

Before Change

        X_ = to_time_series_dataset(X)
        self._norms = numpy.linalg.norm(X_, axis=(1, 2))
        X_ = TimeSeriesScalerMeanVariance(mu=0., std=1.).fit_transform(X_)
        assert X_.shape[-1] == 1, "kShape is supposed to work on monomodal data, provided data has dimension %d" % \
                                  X_.shape[-1]
        if initial_guess is not None:
            assert len(initial_guess) == self.n_clusters, "Initial guess index array must contain {}, {} given".format(self.n_clusters, len(initial_guess))

After Change

        assert X_.shape[-1] == 1, "kShape is supposed to work on monomodal data, provided data has dimension %d" % \
                                  X_.shape[-1]
        if initial_centroids is not None:
            assert initial_centroids.shape[-1] == 1, "kShape is supposed to work on monomodal data, provided data has dimension %d" % \
                                      X_.shape[-1]
            assert initial_centroids.shape[0] == self.n_clusters, "Initial guess index array must contain {}, {} given".format(self.n_clusters, initial_centroids.shape[0])

        rs = check_random_state(self.random_state)

Instances