Skip to content


Subversion checkout URL

You can clone with
Download ZIP


AssertionError: Arrays are not almost equal during test of a fresh install (0.14.1) #2915

mdespriee opened this Issue · 3 comments

4 participants


The test suite fails with this errors (see below)
Any help will be appreciated !!

ERROR: test suite
Traceback (most recent call last):
  File "/usr/lib/python2.6/site-packages/nose/", line 154, in run
  File "/usr/lib/python2.6/site-packages/nose/", line 180, in setUp
    if not self:
  File "/usr/lib/python2.6/site-packages/nose/", line 65, in __nonzero__
    test =
  File "/usr/lib/python2.6/site-packages/nose/", line 223, in generate
    test_func, arg = (test[0], test[1:])
TypeError: 'function' object is unsubscriptable

FAIL: sklearn.linear_model.tests.test_least_angle.test_rank_deficient_design
Traceback (most recent call last):
  File "/usr/lib/python2.6/site-packages/nose/", line 182, in runTest
  File "/usr/lib64/python2.6/site-packages/sklearn/linear_model/tests/", line 218, in test_rank_deficient_design
    assert_array_almost_equal(obj_lars, obj_cd)
  File "/usr/lib64/python2.6/site-packages/numpy/testing/", line 765, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/usr/lib64/python2.6/site-packages/numpy/testing/", line 609, in assert_array_compare
    raise AssertionError(msg)
Arrays are not almost equal

(mismatch 100.0%)
 x: array(5.2333333333333343)
 y: array(0.57115008595591465)

Ran 1697 tests in 91.731s

FAILED (SKIP=15, errors=1, failures=1)
@arjoly arjoly added the Bug label

What is your platform? What is your numpy, scipy version? What is your blas and lapack?


I ran into the same issue with scikit-learn 0.14.1 and 0.15.1. I'm on rhel 6 and using software collections to get python 2.7.5, numpy 1.7.1, scipy 0.12.1, blas 3.2.1, lapack 3.2.1.

How critical is it to get this test to pass?


@johnrc sorry for the late reply, not critical at all, but quite odd.

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.