Broken numpy.eigh

numpy-accelerate.log

Eigenvalues OK for matrix of size 559? (Remove head) True
Eigenvalues OK for matrix of size 559? (Remove tail) True
Eigenvalues OK for matrix of size 560? False

Numpy configuration
version: 1.10.2
lapack_opt_info:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    extra_compile_args = ['-msse3']
    define_macros = [('NO_ATLAS_INFO', 3), ('HAVE_CBLAS', None)]
openblas_lapack_info:
  NOT AVAILABLE
atlas_3_10_blas_threads_info:
  NOT AVAILABLE
atlas_threads_info:
  NOT AVAILABLE
atlas_3_10_threads_info:
  NOT AVAILABLE
atlas_blas_info:
  NOT AVAILABLE
atlas_3_10_blas_info:
  NOT AVAILABLE
atlas_blas_threads_info:
  NOT AVAILABLE
openblas_info:
  NOT AVAILABLE
blas_mkl_info:
  NOT AVAILABLE
blas_opt_info:
    extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
    extra_compile_args = ['-msse3', '-I/System/Library/Frameworks/vecLib.framework/Headers']
    define_macros = [('NO_ATLAS_INFO', 3), ('HAVE_CBLAS', None)]
atlas_info:
  NOT AVAILABLE
atlas_3_10_info:
  NOT AVAILABLE
lapack_mkl_info:
  NOT AVAILABLE
mkl_info:
  NOT AVAILABLE

numpy-mkl.log

$ ~/anaconda2/bin/python testcase.py 
Vendor:  Continuum Analytics, Inc.
Package: mkl
Message: trial mode expires in 30 days
Eigenvalues OK for matrix of size 559? (Remove head) True
Eigenvalues OK for matrix of size 559? (Remove tail) True
Eigenvalues OK for matrix of size 560? True

Numpy configuration
version: 1.10.2
lapack_opt_info:
    libraries = ['mkl_lapack95_lp64', 'mkl_intel_lp64', 'mkl_intel_thread', 'mkl_core', 'iomp5', 'pthread']
    library_dirs = ['/Users/ihaque/anaconda2/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/Users/ihaque/anaconda2/include']
blas_opt_info:
    libraries = ['mkl_intel_lp64', 'mkl_intel_thread', 'mkl_core', 'iomp5', 'pthread']
    library_dirs = ['/Users/ihaque/anaconda2/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/Users/ihaque/anaconda2/include']
openblas_lapack_info:
  NOT AVAILABLE
lapack_mkl_info:
    libraries = ['mkl_lapack95_lp64', 'mkl_intel_lp64', 'mkl_intel_thread', 'mkl_core', 'iomp5', 'pthread']
    library_dirs = ['/Users/ihaque/anaconda2/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/Users/ihaque/anaconda2/include']
blas_mkl_info:
    libraries = ['mkl_intel_lp64', 'mkl_intel_thread', 'mkl_core', 'iomp5', 'pthread']
    library_dirs = ['/Users/ihaque/anaconda2/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/Users/ihaque/anaconda2/include']
mkl_info:
    libraries = ['mkl_intel_lp64', 'mkl_intel_thread', 'mkl_core', 'iomp5', 'pthread']
    library_dirs = ['/Users/ihaque/anaconda2/lib']
    define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
    include_dirs = ['/Users/ihaque/anaconda2/include']

testcase.py

import numpy as np

def compare(mat):
  def eigvals(mat, fn):
    w, v = fn(mat)
    inds = np.argsort(w)
    w = w[inds]
    return w
  w = eigvals(mat, np.linalg.eig)
  wh = eigvals(mat, np.linalg.eigh)
  return np.allclose(w, wh)

rmat = np.asmatrix(np.random.rand(560, 560))
rmat = rmat * rmat.T  # Ensure symmetric

print "Eigenvalues OK for matrix of size 559? (Remove head)", compare(rmat[1:, 1:])
print "Eigenvalues OK for matrix of size 559? (Remove tail)", compare(rmat[:-1, :-1])
print "Eigenvalues OK for matrix of size 560?", compare(rmat)
print
print "Numpy configuration"
print "version:", np.__version__
np.__config__.show()