比较与torch.blackman_window的功能差异

torch.blackman_window

torch.blackman_window(
    window_length,
    periodic=True,
    *,
    dtype=None,
    layout=torch.strided,
    device=None,
    requires_grad=False
) -> Tensor

更多内容详见torch.blackman_window。

mindspore.ops.blackman_window

mindspore.ops.blackman_window(
    window_length,
    periodic=True,
    *,
    dtype=mstype.float32
) -> Tensor

更多内容详见mindspore.ops.blackman_window。

差异对比

PyTorch：返回size与window_length相同的布莱克曼窗，periodic参数确定返回窗口是否会删除对称窗口的最后一个重复值。

MindSpore：MindSpore此API实现功能与PyTorch基本一致，精度稍有差异。

分类	子类	PyTorch	MindSpore	差异
参数	参数1	window_length	window_length	PyTorch中为int类型，MindSpore中为Tensor
	参数2	periodic	periodic	-
	参数3	dtype	dtype	-
	参数4	layout	-	不涉及
	参数5	device	-	不涉及
	参数6	requires_grad	-	MindSpore无此参数，默认支持反向求导

代码示例1

# PyTorch
import torch

torch_output = torch.blackman_window(12, periodic=True)
print(torch_output.numpy())
# [-2.9802322e-08  2.6987284e-02  1.3000000e-01  3.4000000e-01
#   6.3000000e-01  8.9301264e-01  1.0000000e+00  8.9301258e-01
#   6.2999994e-01  3.3999997e-01  1.3000003e-01  2.6987225e-02]

# MindSpore
import mindspore
from mindspore import Tensor

window_length = Tensor(12, mindspore.int32)
ms_output = mindspore.ops.blackman_window(window_length, periodic=True)
print(ms_output.asnumpy())
# [-2.9802322e-08  2.6987284e-02  1.3000000e-01  3.4000000e-01
#   6.3000000e-01  8.9301276e-01  1.0000000e+00  8.9301258e-01
#   6.2999994e-01  3.3999997e-01  1.2999988e-01  2.6987255e-02]