Singular Value Decomposion - Ruby's GSL vs Python's NumPy

gistfile1.rb

# in Ruby with GSL
>> m
=> GSL::Matrix
[  1.000e+00 -1.000e+00 
   1.000e+00  0.000e+00 
   1.000e+00  1.000e+00 ]
>> m.svd
=> [GSL::Linalg::SV::UMatrix
[ -5.774e-01  7.071e-01 
  -5.774e-01  0.000e+00 
  -5.774e-01 -7.071e-01 ], GSL::Linalg::SV::VMatrix
[ -1.000e+00  -0.000e+00 
   -0.000e+00 -1.000e+00 ], GSL::Linalg::SV::SingularValues
[ 1.732e+00 1.414e+00 ]]

# in Python with numpy
>>> y
matrix([[ 1, -1],
        [ 1,  0],
        [ 1,  1]])
>>> np.linalg.svd(y)
(matrix([[ -5.77350269e-01,   7.07106781e-01,   4.08248290e-01],
        [ -5.77350269e-01,   5.55111512e-17,  -8.16496581e-01],
        [ -5.77350269e-01,  -7.07106781e-01,   4.08248290e-01]]), array([ 1.73205081,  1.41421356]), matrix([[-1., -0.],
        [-0., -1.]]))