mindspore.nn.MatInverse

class mindspore.nn.MatInverse[source]

Calculates the inverse of Positive-Definite Hermitian matrix using Cholesky decomposition.

Inputs:

a (Tensor[Number]) - The input tensor. It must be a positive-definite matrix. With float16 or float32 data type.

Outputs:

Tensor, has the same dtype as the a.

Raises: TypeError – If dtype of a is neither float16 nor float32.

Supported Platforms:: GPU

Examples

>>> input_a = Tensor(np.array([[4, 12, -16], [12, 37, -43], [-16, -43, 98]]).astype(np.float32))
>>> op = nn.MatInverse()
>>> output = op(input_a)
>>> print(output)
[[49.36112  -13.555558  2.1111116]
 [-13.555558  3.7777784  -0.5555557]
 [2.1111116  -0.5555557  0.11111111]]