mindspore.mint.addmv

mindspore.mint.addmv(input, mat, vec, *, beta=1, alpha=1)

Performs a matrix-vector product of mat and vec, and add the input vector input to the final result.

If mat is a tensor of size \((N, M)\) , vec is a 1-D tensor of size \(M\) , then input must be broadcastable with a 1-D tensor of size \(N\) . In this case, output is a 1-D Tensor of size \(N\) .

\[output = eta input + lpha (mat @ vec)\]

Warning

This is an experimental API that is subject to change or deletion.

Parameters

• input (Tensor) – Vector to be added.

• mat (Tensor) – The first tensor needs to be multiplied.

• vec (Tensor) – The second tensor needs to be multiplied.

Keyword Arguments

• beta (Union[float, int], optional) – Coefficient of input. Default: 1.

• alpha (Union[float, int], optional) – Coefficient of \(mat @ vec\) . Default: 1.

Returns

Tensor, with a shape of \((N,)\) , and its dtype is the same as input.

Raises

• TypeError – If dtype of input, mat or vec is not tensor.

• TypeError – If dtypes of mat and vec are not the same.

• ValueError – If mat is not a 2-D tensor.

• ValueError – If vec is not a 1-D tensor.

Supported Platforms:

Ascend

Examples

>>> import numpy as np
>>> from mindspore import Tensor, mint
>>> input = Tensor(np.array([2., 3.]).astype(np.float32))
>>> mat = Tensor(np.array([[2., 5., 3.], [4., 2., 2.]]).astype(np.float32))
>>> vec = Tensor(np.array([3., 2., 4.]).astype(np.float32))
>>> output = mint.addmv(input, mat, vec)
>>> print(output)
[30. 27.]