cr.sparse.la.subspaces.smallest_principal_angles_cos

cr.sparse.la.subspaces.smallest_principal_angles_cos(subspaces)

Returns the smallest principal angles between each pair of subspaces

Parameters

A (jax.numpy.ndarray) – An array of ONBs for the subspaces

Returns

A symmetric matrix containing the cosine of the

smallest principal angles between each pair of subspaces

Return type

(jax.numpy.ndarray)

Further reading on implementation:

• Vectorizing computations on pairs of elements in an nd-array

• SO: How to vectorize a 2 level loop in NumPy