from sklearn.datasets import load_iris

from sklearn.cluster import Birch

​

X, y = load_iris(return_X_y=True, as_frame=True)

birch = Birch(n_clusters=3)

birch.fit_predict(X)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Input In [4], in <cell line: 6>()
      4 X, y = load_iris(return_X_y=True, as_frame=True)
      5 birch = Birch(n_clusters=3)
----> 6 birch.fit_predict(X)

File ~/Documents/packages/scikit-learn/sklearn/base.py:716, in ClusterMixin.fit_predict(self, X, y)
    698 """
    699 Perform clustering on `X` and returns cluster labels.
    700 
   (...)
    712     Cluster labels.
    713 """
    714 # non-optimized default implementation; override when a better
    715 # method is possible for a given clustering algorithm
--> 716 self.fit(X)
    717 return self.labels_

File ~/Documents/packages/scikit-learn/sklearn/cluster/_birch.py:545, in Birch.fit(self, X, y)
    543 # TODO: Remove deprecated flags in 1.2
    544 self._deprecated_fit, self._deprecated_partial_fit = True, False
--> 545 return self._fit(X, partial=False)

File ~/Documents/packages/scikit-learn/sklearn/cluster/_birch.py:615, in Birch._fit(self, X, partial)
    612 self.subcluster_centers_ = centroids
    613 self._n_features_out = self.subcluster_centers_.shape[0]
--> 615 self._global_clustering(X)
    616 return self

File ~/Documents/packages/scikit-learn/sklearn/cluster/_birch.py:762, in Birch._global_clustering(self, X)
    759     self.subcluster_labels_ = clusterer.fit_predict(self.subcluster_centers_)
    761 if compute_labels:
--> 762     self.labels_ = self._predict(X)

File ~/Documents/packages/scikit-learn/sklearn/cluster/_birch.py:700, in Birch._predict(self, X)
    697 kwargs = {"Y_norm_squared": self._subcluster_norms}
    699 with config_context(assume_finite=True):
--> 700     argmin = pairwise_distances_argmin(
    701         X, self.subcluster_centers_, metric_kwargs=kwargs
    702     )
    703 return self.subcluster_labels_[argmin]

File ~/Documents/packages/scikit-learn/sklearn/metrics/pairwise.py:791, in pairwise_distances_argmin(X, Y, axis, metric, metric_kwargs)
    788         metric = "sqeuclidean"
    789         metric_kwargs = {}
--> 791     indices = PairwiseDistancesArgKmin.compute(
    792         X=X,
    793         Y=Y,
    794         k=1,
    795         metric=metric,
    796         metric_kwargs=metric_kwargs,
    797         strategy="auto",
    798         return_distance=False,
    799     )
    800     indices = indices.flatten()
    801 else:
    802     # TODO: once PairwiseDistancesArgKmin supports sparse input matrices and 32 bit,
    803     # we won't need to fallback to pairwise_distances_chunked anymore.
    804 
    805     # Turn off check for finiteness because this is costly and because arrays
    806     # have already been validated.

File ~/Documents/packages/scikit-learn/sklearn/metrics/_pairwise_distances_reduction/_dispatcher.py:233, in PairwiseDistancesArgKmin.compute(cls, X, Y, k, metric, chunk_size, metric_kwargs, strategy, return_distance)
    227 # Note (jjerphan): Some design thoughts for future extensions.
    228 # This factory comes to handle specialisations for the given arguments.
    229 # For future work, this might can be an entrypoint to specialise operations
    230 # for various backend and/or hardware and/or datatypes, and/or fused
    231 # {sparse, dense}-datasetspair etc.
    232 if X.dtype == Y.dtype == np.float64:
--> 233     return PairwiseDistancesArgKmin64.compute(
    234         X=X,
    235         Y=Y,
    236         k=k,
    237         metric=metric,
    238         chunk_size=chunk_size,
    239         metric_kwargs=metric_kwargs,
    240         strategy=strategy,
    241         return_distance=return_distance,
    242     )
    243 raise ValueError(
    244     "Only 64bit float datasets are supported at this time, "
    245     f"got: X.dtype={X.dtype} and Y.dtype={Y.dtype}."
    246 )

File ~/Documents/packages/scikit-learn/sklearn/metrics/_pairwise_distances_reduction/_argkmin.pyx:74, in sklearn.metrics._pairwise_distances_reduction._argkmin.PairwiseDistancesArgKmin64.compute()

File ~/Documents/packages/scikit-learn/sklearn/metrics/_pairwise_distances_reduction/_argkmin.pyx:343, in sklearn.metrics._pairwise_distances_reduction._argkmin.FastEuclideanPairwiseDistancesArgKmin64.__init__()

File ~/Documents/packages/scikit-learn/sklearn/metrics/_pairwise_distances_reduction/_datasets_pair.pyx:98, in sklearn.metrics._pairwise_distances_reduction._datasets_pair.DatasetsPair.get_for()

File ~/Documents/packages/scikit-learn/sklearn/metrics/_pairwise_distances_reduction/_datasets_pair.pyx:146, in sklearn.metrics._pairwise_distances_reduction._datasets_pair.DenseDenseDatasetsPair.__init__()

File stringsource:658, in View.MemoryView.memoryview_cwrapper()

File stringsource:349, in View.MemoryView.memoryview.__cinit__()

ValueError: ndarray is not C-contiguous