scipy.sparse.linalg.inv cannot take ndarray as argument despite what documentation says

[gdupras]

The documentation (scipy.sparse.linalg.inv) says

Parameters

A(M,M) ndarray or sparse matrix

square matrix to be inverted

But this code

import numpy as np
from scipy.sparse import linalg
linalg.inv(np.array([[1, 2], [3, 4]]))

throws:

TypeError: Input must be a sparse matrix

[WarrenWeckesser]

I suspect the best option here is to fix the docs.

[gdupras]

Would it be possible for scipy.sparse.linalg.inv to call np.linalg.inv if A is an ndarray?

[WarrenWeckesser]

@gdupras, yes, that's possible, and it may have been the original intent of the docstring author, but the explicit check for a sparse input has been there for many years, so I suspect there isn't much demand for that convenience. @perimosocordiae, @pv: any preferences?

[stefanv]

May be best to fix the docs here, since it is not clear which inv to dispatch to otherwise (scipy? numpy?).