check_2D_num_array

pybear.base.check_2D_num_array(X, require_all_finite=False)

Validate things that are expected to be 2D arrays of numbers.

Accepts 2D Python built-ins, numpy arrays, pandas dataframes, polars dataframes, and all scipy sparse matrices/arrays. Python built-ins can be ragged. When require_all_finite is True, every element in the array must be an instance of numbers.Number; a ValueError will be raised if there are any nan-like or infinity-like values. If require_all_finite is False, non-finite values are ignored and only the finite values must be an instance of numbers.Number. If all checks pass then None is returned.

Parameters:

XXContainer[numbers.Number]

Something that is expected to be a 2D array of numbers.

require_all_finitebool, default=False

If True, disallow all non-finite values, such as nan-like or infinity-like values.

Returns:

None

Raises:

TypeError:

For invalid container.

ValueError:

For non-finite values when require_all_finite is True.

Notes

Type Aliases

Python2DTypes:

list[list] | tuple[tuple]

Numpy2DTypes:

numpy.ndarray

Pandas2DTypes:

pandas.DataFrame

Polars2DTypes:

polars.DataFrame

ScipySparseTypes:

ss.csc_matrix | ss.csc_array | ss.csr_matrix | ss.csr_array | ss.coo_matrix | ss.coo_array | ss.dia_matrix | ss.dia_array | ss.lil_matrix | ss.lil_array | ss.dok_matrix | ss.dok_array | ss.bsr_matrix | ss.bsr_array

XContainer:

Python2DTypes | Numpy2DTypes | Pandas2DTypes | Polars2DTypes | ScipySparseTypes

Examples

>>> from pybear.base import check_2D_num_array
>>> import numpy as np
>>> X = np.random.randint(0, 10, (37, 13)).astype(np.float64)
>>> X[0][8] = np.nan
>>> X[31][3] = np.inf
>>> check_2D_num_array(X, require_all_finite=False)
>>> try:
...     check_2D_num_array(X, require_all_finite=True)
... except ValueError as e:
...     print(e)
Got non-finite values when not allowed.