mindspore.Tensor.diag

Tensor.diag()[source]

Constructs a diagonal tensor with a given diagonal values.

Assume self tensor has dimensions \([D_1,... D_k]\), the output is a tensor of rank 2k with dimensions \([D_1,..., D_k, D_1,..., D_k]\) where: \(output[i_1,..., i_k, i_1,..., i_k] = self[i_1,..., i_k]\) and 0 everywhere else.

Returns

Tensor, has the same dtype as self tensor.

Raises

ValueError – If rank of self tensor is less than 1.

Supported Platforms:

Ascend GPU

Examples

>>> from mindspore import Tensor
>>> x = Tensor([1, 2, 3, 4]).astype('int32')
>>> output = x.diag()
>>> print(output)
[[1 0 0 0]
 [0 2 0 0]
 [0 0 3 0]
 [0 0 0 4]]