比较与torch.meshgrid的差异
torch.meshgrid
torch.meshgrid(
*tensors)
更多内容详见torch.meshgrid。
mindspore.ops.meshgrid
mindspore.ops.meshgrid(*inputs, indexing='xy')
更多内容详见mindspore.ops.meshgrid。
差异对比
PyTorch:从给定的tensors生成网格矩阵。tensors如果是scalar的list,则scalar将自动被视为大小为(1,)的张量。
MindSpore:MindSpore此API实现功能与PyTorch一致。inputs参数只支持Tensor,不支持scalar。
分类 |
子类 |
PyTorch |
MindSpore |
差异 |
---|---|---|---|---|
参数 |
参数1 |
tensors |
inputs |
功能一致 |
参数2 |
- |
indexing |
torch.meshgrid v1.8.1无参数 |
代码示例1
# PyTorch
import numpy as np
import torch
x = torch.tensor(np.array([1, 2, 3, 4]).astype(np.int32))
y = torch.tensor(np.array([5, 6, 7]).astype(np.int32))
z = torch.tensor(np.array([8, 9, 0, 1, 2]).astype(np.int32))
output = torch.meshgrid(x, y, z)
print(output)
# (tensor([[[1, 1, 1, 1, 1],
# [1, 1, 1, 1, 1],
# [1, 1, 1, 1, 1]],
# [[2, 2, 2, 2, 2],
# [2, 2, 2, 2, 2],
# [2, 2, 2, 2, 2]],
# [[3, 3, 3, 3, 3],
# [3, 3, 3, 3, 3],
# [3, 3, 3, 3, 3]],
# [[4, 4, 4, 4, 4],
# [4, 4, 4, 4, 4],
# [4, 4, 4, 4, 4]]], dtype=torch.int32), tensor([[[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]]], dtype=torch.int32), tensor([[[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]]], dtype=torch.int32))
# MindSpore
import mindspore
import numpy as np
from mindspore import Tensor
x = Tensor(np.array([1, 2, 3, 4]).astype(np.int32))
y = Tensor(np.array([5, 6, 7]).astype(np.int32))
z = Tensor(np.array([8, 9, 0, 1, 2]).astype(np.int32))
output = mindspore.ops.meshgrid(x, y, z, indexing='ij')
print(output)
# (Tensor(shape=[4, 3, 5], dtype=Int32, value=
# [[[1, 1, 1, 1, 1],
# [1, 1, 1, 1, 1],
# [1, 1, 1, 1, 1]],
# [[2, 2, 2, 2, 2],
# [2, 2, 2, 2, 2],
# [2, 2, 2, 2, 2]],
# [[3, 3, 3, 3, 3],
# [3, 3, 3, 3, 3],
# [3, 3, 3, 3, 3]],
# [[4, 4, 4, 4, 4],
# [4, 4, 4, 4, 4],
# [4, 4, 4, 4, 4]]]), Tensor(shape=[4, 3, 5], dtype=Int32, value=
# [[[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]],
# [[5, 5, 5, 5, 5],
# [6, 6, 6, 6, 6],
# [7, 7, 7, 7, 7]]]), Tensor(shape=[4, 3, 5], dtype=Int32, value=
# [[[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]],
# [[8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2],
# [8, 9, 0, 1, 2]]]))