mindscience.e3nn.o3.Norm

class mindscience.e3nn.o3.Norm(irreps_in, squared=False, dtype=mindspore.float32, ncon_dtype=mindspore.float32)[source]

Compute the norm (length) of each irreducible representation (irrep) contained in a direct-sum tensor.

Given a tensor that transforms under a direct sum of irreps, this module returns a tensor whose entries are the norms of the individual irreps. For example, if the input contains three vectors (3x1o irreps), the output will be a 3-component tensor where each entry is the Euclidean norm of the corresponding vector.

Parameters

irreps_in (Union[str, Irrep, Irreps]) – Irreps for the input.
squared (bool, optional) – Whether to return the squared norm. Default: False.
dtype (mindspore.dtype, optional) – The type of input tensor. Default: mindspore.float32.
ncon_dtype (mindspore.dtype, optional) – The type of input tensors of ncon computation module. Default: mindspore.float32.

Inputs:

v (Tensor) - The shape of Tensor is \((..., irreps\_in.dim)\) .

Outputs:

output (Tensor) - The shape of Tensor is \((..., irreps\_out.dim)\) .

Examples

>>> import mindspore as ms
>>> import numpy as np
>>> from mindscience.e3nn.o3 import Norm
>>> n = Norm('3x1o')
>>> v = ms.Tensor(np.linspace(1., 2., n.irreps_in.dim), dtype=ms.float32)
>>> n(v).shape
(1, 3)