scipy.sparse.

expand_dims#

scipy.sparse.expand_dims(A, /, *, axis=0)[source]#

Add trivial axes to an array. Shape gets a 1 inserted at position axis.

Parameters:

Asparse array
axisint: Position in the expanded axes where the new axis (or axes) is placed. For a dimension N array, a valid axis is an integer on the closed-interval [-N-1, N]. Negative values work from the end of the shape. 0 prepends an axis, as does -N-1. -1 appends an axis, as does N. The new axis has shape 1 and indices are created with the value 0.

Returns:

outsparse array: A expanded copy output in COO format with the same dtype as A.

Raises:

ValueError: If provided a non-integer or out of range [-N-1, N] axis, where N is A.ndim.

Examples

>>> from scipy.sparse import csr_array, expand_dims
>>> A = csr_array([[1, 2], [2, 0]])
>>> A.shape
(2, 2)
>>> expand_dims(A, axis=1).shape
(2, 1, 2)