ImportError: cannot import name check_array from sklearn.utils.validation

Question

When I import the function check_array from module sklearn.utils.validation , it got an Import Error ( ImportError: cannot import name check_array ). The tab completion got check_arrays , but I'm wondering there only exists a function called check_array in validation.py ( source code on Github ). Besides, the spectral clustering algorithm implemented in scikit-learn/sklearn/cluster/spectral.py also used from ..utils.validation import check_array , not check_arrays . I'm quite confused about this, and my scikit-learn version is '0.15.0b1' . Hope somebody gives me a clue.

sample code

import numpy as np from sklearn.utils.validation import check_array def my_fit_affinity(X, affinity_type, n_neighbors, **kernel_params): """ Create an affinity matrix for X using the selected affinity type """ X = check_array(X, accept_sparse = ['csr', 'csc', 'coo']) .... return affinity_matrix_

Answer 1

如果您使用 jupyter notebook，请尝试重新启动内核。

Answer 2

To those searching, as @eickenberg mentioned in the comments, this was something that was changed between versions. Using the latest version of scikit-learn resolved the issue.

Answer 3

点击内核>重新启动并运行所有