BUG: linalg.subspace_angles gives wrong results

[c-meyer]

The function subspace_angles gives wrong results this can be shown with a simple example:
Two matrices have two vectors pointing to the same direction, but the last ones are orthogonoal to each other and are not contained in the space of the other ones. Therefore the subspace angles must be 0, 0 and pi/2.

Reproducing code example:

import numpy as np
from scipy.linalg import subspace_angles

A = np.array([[1, 0, 0],
       [0, 1, 0],
       [0, 0, 1],
       [0, 0, 0],
       [0, 0, 0]])

B = np.array([[1. , 0. , 0. ],
       [0. , 1. , 0. ],
       [0. , 0. , 0. ],
       [0. , 0., 0. ],
       [0. , 0. , 1. ]])

sangles = subspace_angles(A,B)
print(sangles) # Produces wrong angles!

Error message:

# No error message

Scipy/Numpy/Python version information:

<<Output from 'import sys, scipy, numpy; print(scipy.__version__, numpy.__version__, sys.version_info)'>>
1.0.0 1.14.0 sys.version_info(major=3, minor=6, micro=4, releaselevel='final', serial=0)

[c-meyer]

I am going to send a pull request for this issue.