pytorch如何增加维度_PyTorch ---- torch.nn.function.pad 函数用法（补充维度上的数值）...

1.二维数组：对最内部元素左侧增加元素(例如 1 的左侧)

a = torch.tensor([[1, 2, 3, 4], [1, 2, 3, 4]])

a1 = torch.nn.functional.pad(a, pad=(1, 0, 0, 0), mode='constant', value=1)

print("a = ", a)

print("a1 = ", a1)

运行结果：

a = tensor([[1, 2, 3, 4],

[1, 2, 3, 4]])

a1 = tensor([[1, 1, 2, 3, 4],

[1, 1, 2, 3, 4]])

2.二维数组：对最内部元素右侧增加元素(例如 4 右侧)

a = torch.tensor([[1, 2, 3, 4], [1, 2, 3, 4]])

a1 = torch.nn.functional.pad(a, pad=(0, 1, 0, 0), mode='constant', value=1)

print("a = ", a)

print("a1 = ", a1)

运行结果：

a = tensor([[1, 2, 3, 4],

[1, 2, 3, 4]])

a1 = tensor([[1, 2, 3, 4, 1],

[1, 2, 3, 4, 1]])

3.二维数组：对最内部一维数组左侧增加元素(例如 [1, 2, 3, 4] 左侧)

a = torch.tensor([[1, 2, 3, 4], [1, 2, 3, 4]])

a1 = torch.nn.functional.pad(a, pad=(0, 0, 1, 0), mode='constant', value=1)

print("a = ", a)

print("a1 = ", a1)

运行结果：

a = tensor([[1, 2, 3, 4],

[1, 2, 3, 4]])

a1 = tensor([[1, 1, 1, 1],

[1, 2, 3, 4],

[1, 2, 3, 4]])

4.二维数组：对最内部一维数组右