Hi Dawen,
I can confirm the error but its origin is not in sklearn but in numpy
(current dev version). Please try the following::
import numpy as np
from scipy import sparse as sp
A = np.random.rand(10, 10)
S = sp.csr_matrix(A)
_ = np.dot(A, A) # this works OK
_ = np.dot(S, S) # this segfaults!
I'll check with the numpy folks if that's intended behaviour - if so,
we have to change our sparse matrix handling....
I'll keep you posted.
thanks,
Peter
2011/11/9 Dawen Liang
<reply+i-2181668-b6429fd02d66f2dbe58d9f3c3df3b590358f9...@reply.github.com>:
> Hi all,
>
> Thanks to the help this afternoon, I finally got sklearn working with my test
> logistic regression. However, while I run the test, I got the segmentation
> fault at last, here is the output (very long), it seems to me that something
> is not implemented yet and not sure if that's the cause of the error:
>
>
> Running unit tests and doctests for sklearn
> NumPy version 2.0.0.dev-7297785
> NumPy is installed in
> /Library/Python/2.7/site-packages/numpy-2.0.0.dev_7297785_20111104-py2.7-macosx-10.7-x86_64.egg/numpy
> Python version 2.7.1 (r271:86832, Jun 16 2011, 16:59:05) [GCC 4.2.1 (Based on
> Apple Inc. build 5658) (LLVM build 2335.15.00)]
> nose version 1.1.2
> /Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/cross_val.py:2:
> UserWarning: sklearn.cross_val namespace is deprecated in version 0.9 and
> will be removed in version 0.11. Please use sklearn.cross_validation instead.
> warnings.warn('sklearn.cross_val namespace is deprecated in version 0.9'
> /Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/hmm.py:16:
> UserWarning: sklearn.hmm is orphaned, undocumented and has known numerical
> stability issues. If nobody volunteers to write documentation and make it
> more stable, this module will be removed in version 0.11.
> warnings.warn('sklearn.hmm is orphaned, undocumented and has known numerical'
> .........../Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/neighbors/base.py:23:
> UserWarning: kneighbors: neighbor k+1 and neighbor k have the same distance:
> results will be dependent on data order.
> warnings.warn(msg)
> ......../Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/cluster/k_means_.py:219:
> UserWarning: Explicit initial center position passed: performing only one
> init in the k-means
> warnings.warn('Explicit initial center position passed: '
> .............../Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/cluster/spectral.py:77:
> UserWarning: pyamg not available, using scipy.sparse
> warnings.warn('pyamg not available, using scipy.sparse')
> .S...................S.S.........................FFF............/Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/decomposition/fastica_.py:199:
> UserWarning: Please note: the interface of fastica has changed: X is now
> assumed to be of shape [n_samples, n_features]
> warnings.warn("Please note: the interface of fastica has changed: "
> ..................../Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/decomposition/nmf.py:237:
> UserWarning: Iteration limit reached in nls subproblem.
> warnings.warn("Iteration limit reached in nls subproblem.")
> .................../Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/decomposition/sparse_pca.py:146:
> RuntimeWarning: invalid value encountered in divide
> U /= np.sqrt((U ** 2).sum(axis=0))
> ......................./Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/externals/joblib/test/test_func_inspect.py:101:
> UserWarning: Cannot inspect object <functools.partial object at
> 0x103a760a8>, ignore list will not work.
> nose.tools.assert_equal(filter_args(ff, ['y'], 1),
> ........................................................................................________________________________________________________________________________
> test_memory setup
> ________________________________________________________________________________
> ________________________________________________________________________________
> test_memory teardown
> ________________________________________________________________________________
> ..________________________________________________________________________________
> test_memory setup
> ________________________________________________________________________________
> ......................................................................................................................________________________________________________________________________________
> test_memory teardown
> ________________________________________________________________________________
> .________________________________________________________________________________
> setup numpy_pickle
> ________________________________________________________________________________
> .......................................Failed to save <class
> 'numpy.ma.core.MaskedArray'> to .npy file:
> Traceback (most recent call last):
> File
> "/Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/externals/joblib/numpy_pickle.py",
> line 62, in save
> self.np.save(filename, obj)
> File
> "/Library/Python/2.7/site-packages/numpy-2.0.0.dev_7297785_20111104-py2.7-macosx-10.7-x86_64.egg/numpy/lib/npyio.py",
> line 434, in save
> format.write_array(fid, arr)
> File
> "/Library/Python/2.7/site-packages/numpy-2.0.0.dev_7297785_20111104-py2.7-macosx-10.7-x86_64.egg/numpy/lib/format.py",
> line 409, in write_array
> array.tofile(fp)
> File
> "/Library/Python/2.7/site-packages/numpy-2.0.0.dev_7297785_20111104-py2.7-macosx-10.7-x86_64.egg/numpy/ma/core.py",
> line 5380, in tofile
> raise NotImplementedError("Not implemented yet, sorry...")
> NotImplementedError: Not implemented yet, sorry...
>
> .________________________________________________________________________________
> teardown numpy_pickle
> ________________________________________________________________________________
> ...............E..................................F................/Library/Python/2.7/site-packages/scikit_learn-0.10_git-py2.7-macosx-10.7-intel.egg/sklearn/linear_model/least_angle.py:218:
> RuntimeWarning: invalid value encountered in divide
> z = -coefs[n_iter, active] / least_squares
> ....Segmentation fault: 11
>
>
> Any help will be appreciated. Thanks!
>
> ---
> Reply to this email directly or view it on GitHub:
> https://github.com/scikit-learn/scikit-learn/issues/435
>
--
Peter Prettenhofer
------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
