Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

Loading…

installing on 64bit windows 7 for Python 2.7 using unofficial installers #2065

Closed
etseidler opened this Issue · 6 comments

4 participants

@etseidler

So I took your suggestion and installed numpy, scipy, matplotlib and then scikit-learn (in that order) all from the same website that you listed (http://www.lfd.uci.edu/~gohlke/pythonlibs/#scikit-learn which was very helpful, btw). Out of curiosity, I ran the nose tests and got the output below. I'm pretty sure I can still use most of the scikit-learn functionality, but would anyone care to comment on that? Again, my approach was to use only the installers. Am I overlooking anything?

C:\Python27\scripts>nosetests.exe sklearn --exe
...............................................................C:\Python27\lib\site-packages\sklearn\manifold\spectral_embedding.py:225: UserWarning: Graph is not fully connected, spectral embedding may not works as expected.
  warnings.warn("Graph is not fully connected, spectral embedding"
.......C:\Python27\lib\site-packages\sklearn\covariance\graph_lasso_.py:45: RuntimeWarning: invalid value encountered in absolute
  gap += alpha * (np.abs(precision_).sum()
C:\Python27\lib\site-packages\sklearn\covariance\graph_lasso_.py:46: RuntimeWarning: invalid value encountered in absolute
  - np.abs(np.diag(precision_)).sum())
C:\Python27\lib\site-packages\numpy\linalg\linalg.py:1664: RuntimeWarning: invalid value encountered in absolute
  absd = absolute(d)
C:\Python27\lib\site-packages\sklearn\covariance\graph_lasso_.py:32: RuntimeWarning: invalid value encountered in absolute
  cost += alpha * (np.abs(precision_).sum()
C:\Python27\lib\site-packages\sklearn\covariance\graph_lasso_.py:33: RuntimeWarning: invalid value encountered in absolute
  - np.abs(np.diag(precision_)).sum())
C:\Python27\lib\site-packages\sklearn\covariance\graph_lasso_.py:195: RuntimeWarning: invalid value encountered in absolute
  if np.abs(d_gap) < tol:
....SS......C:\Python27\lib\site-packages\sklearn\datasets\tests\test_base.py:124: UserWarning: Could not load sample images, PIL is not available.
  warnings.warn("Could not load sample images, PIL is not available.")
...C:\Python27\lib\site-packages\sklearn\datasets\tests\test_base.py:145: UserWarning: Could not load sample images, PIL is not available.
  warnings.warn("Could not load sample images, PIL is not available.")
.C:\Python27\lib\site-packages\sklearn\datasets\tests\test_base.py:161: UserWarning: Could not load sample images, PIL is not available.
  warnings.warn("Could not load sample images, PIL is not available.")
.....SS................................................S.........................................................S.........................................SSS......................C:\Python27\lib\site-packages\sklearn\externals\joblib\test
test_func_inspect.py:122: UserWarning: Cannot inspect object <functools.partial object at 0x000000000F7D7E08>, ignore list will not work.
  nose.tools.assert_equal(filter_args(ff, ['y'], (1, )),
...............................................................................................................................................................................................................................................
...............................................................................................................................................................................................................................................
...............................................An unexpected error occurred while tokenizing input
The following traceback may be corrupted or invalid
The error message is: ('EOF in multi-line statement', (63, 0))

An unexpected error occurred while tokenizing input
The following traceback may be corrupted or invalid
The error message is: ('EOF in multi-line statement', (63, 0))

....S........................................................................................................S.F...............................................................................................................................
...............................................................................................................................................................................................................................................
.........................................................................................................................................EE...........................................SSS....S....S............................................
......................................................................................
======================================================================
ERROR: sklearn.tests.test_dummy.test_stratified_strategy
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\nose-1.3.0-py2.7.egg\nose\case.py", line 197, in runTest
    self.test(*self.arg)
  File "C:\Python27\lib\site-packages\sklearn\tests\test_dummy.py", line 96, in test_stratified_strategy
    y_pred = clf.predict(X)
  File "C:\Python27\lib\site-packages\sklearn\dummy.py", line 133, in predict
    proba = self.predict_proba(X)
  File "C:\Python27\lib\site-packages\sklearn\dummy.py", line 197, in predict_proba
    out = rs.multinomial(1, class_prior_[k], size=n_samples)
  File "mtrand.pyx", line 4257, in mtrand.RandomState.multinomial (numpy\random\mtrand\mtrand.c:20210)
TypeError: unsupported operand type(s) for +: 'long' and 'tuple'

======================================================================
ERROR: sklearn.tests.test_dummy.test_stratified_strategy_multioutput
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\nose-1.3.0-py2.7.egg\nose\case.py", line 197, in runTest
    self.test(*self.arg)
  File "C:\Python27\lib\site-packages\sklearn\tests\test_dummy.py", line 115, in test_stratified_strategy_multioutput
    y_pred = clf.predict(X)
  File "C:\Python27\lib\site-packages\sklearn\dummy.py", line 133, in predict
    proba = self.predict_proba(X)
  File "C:\Python27\lib\site-packages\sklearn\dummy.py", line 197, in predict_proba
    out = rs.multinomial(1, class_prior_[k], size=n_samples)
  File "mtrand.pyx", line 4257, in mtrand.RandomState.multinomial (numpy\random\mtrand\mtrand.c:20210)
TypeError: unsupported operand type(s) for +: 'long' and 'tuple'

======================================================================
FAIL: Test BayesianRegression ARD classifier
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Python27\lib\site-packages\nose-1.3.0-py2.7.egg\nose\case.py", line 197, in runTest
    self.test(*self.arg)
  File "C:\Python27\lib\site-packages\sklearn\linear_model\tests\test_bayes.py", line 58, in test_toy_ard_object
    assert(np.abs(clf.predict(test) - [1, 3, 4]).sum() < 1.e-3)  # identity
AssertionError

----------------------------------------------------------------------
Ran 1598 tests in 96.640s

FAILED (SKIP=16, errors=2, failures=1)

C:\Python27\scripts>
@larsmans
Owner

Can you tell us the output from:

import numpy, scipy, sys
print(numpy.__version__)
print(scipy.__version__)
print(sys.version)

?

@etseidler
@larsmans
Owner

I'm guessing it's NumPy 1.7 that's causing the issue. Hold on.

@larsmans
Owner

I'm sorry, but I can't reproduce this with NumPy 1.7.1 and SciPy 0.12.0 on Linux.

@etseidler
@amueller amueller added this to the 0.15.1 milestone
@ogrisel
Owner

This was already fixed. I built scikit-learn 0.15.0 against Christoph Gohlke's binaries for numpy and scipy.

@ogrisel ogrisel closed this
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.