check_scipy_sparse

pybear.base.check_scipy_sparse(X, allowed)

Check whether a passed data container is a scipy sparse matrix / array.

If it is, check the type against the allowed types specified by the user in allowed. If X is not a scipy sparse container, skip all checks and return None. If X is an allowed scipy sparse container, return None. If X is a disallowed scipy container, do not recast the passed container to a valid scipy sparse container but raise a TypeError.

Parameters:

Xarray_like of shape (n_samples, n_features) or (n_samples, )

The data to be checked whether it is an allowed scipy sparse matrix / array. This parameter is not checked for being a valid data container. It only undergoes checks if it is a scipy sparse container.

allowedLiteral[False] | None | Sequence[str]

If None or False, disallow any scipy sparse containers. Otherwise, a vector-like sequence of literals indicating the types of scipy sparse matrices / arrays that are allowed. Valid literals are ‘csr’, ‘csc’, ‘coo’, ‘dia’, ‘lil’, ‘dok’, and ‘bsr’. If a disallowed scipy sparse type is passed it is not recast to a valid type, but a TypeError is raised.

Returns:

None

Raises:

TypeError:

If allowed is None or False and X is a scipy sparse container.

If X is a scipy sparse container but not one of the allowed containers.

Examples

>>> from pybear.base import check_scipy_sparse
>>> import numpy as np
>>> import scipy.sparse as ss
>>> X_np = np.random.uniform(0, 1, (5, 3))
>>> X_csc = ss.csc_array(X_np)
>>> print(check_scipy_sparse(X_csc, ['csc', 'csr', 'coo']))
None
>>> try:
...     check_scipy_sparse(X_csc, False)
... except Exception as e:
...     print(repr(e)[:53])
TypeError("X is <class 'scipy.sparse._csc.csc_array'>