torch.gather

函数声明为：

torch.gather(input, dim, index, out=None, sparse_grad=False) → Tensor

函数作用为：

Gathers values along an axis specified by dim.

\[\begin{array}{l}{\text { If input is an n-dimensional tensor with size }\left(x_{0}, x_{1} \ldots, x_{i-1}, x_{i}, x_{i+1}, \ldots, x_{n-1}\right) \text { and dim }=i, \text { then index must be }} \\ {\text { an } n \text { -dimensional tensor with size }\left(x_{0}, x_{1}, \ldots, x_{i-1}, y, x_{i+1}, \ldots, x_{n-1}\right) \text { where } y \geq 1 \text { and out will have the same }} \\ {\text { size as index. }}\end{array}\]

举个例子：

t = torch.tensor([[1, 2], [3, 4]])
#tensor([[3, 4],
#        [1, 2],
#        [1, 4]])
torch.gather(t, 0, torch.tensor([[1, 1], [0, 0], [0, 1]]))

由于输出张量和index张量有着相同的形状，所以我们可以做以下直观的理解：

torch.unsqueeze

函数声明为：

torch.unsqueeze(input, dim, out=None) → Tensor

函数的作用为：

Returns a new tensor with a dimension of size one inserted at the specified position.

可以理解为使得这个张量在指定的位置增加一维，例子如下：

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

# insert a dimension in dim 0
torch.unsqueeze(x, 0)
# tensor([[ 1,  2,  3,  4]])

# insert a dimension in dim 1
torch.unsqueeze(x, 1)
# tensor([[ 1],
#        [ 2],
#        [ 3],
#        [ 4]])

torch.squeeze

函数声明为：

torch.squeeze(input, dim=None, out=None) → Tensor

函数作用为：

Returns a tensor with all the dimensions of input of size 1 removed.

可以理解为返回一个张量，这个张量所有size为1的维度都将被去除，如果给定了dim，那么我们将会只去除这一维度（如果这一维度size为1）：

>>> x = torch.zeros(2, 1, 2, 1, 2)
>>> x.size()
torch.Size([2, 1, 2, 1, 2])
>>> y = torch.squeeze(x)
>>> y.size()
torch.Size([2, 2, 2])
>>> y = torch.squeeze(x, 0)
>>> y.size()
torch.Size([2, 1, 2, 1, 2])
>>> y = torch.squeeze(x, 1)
>>> y.size()
torch.Size([2, 2, 1, 2])

torch.Tensor.view

函数声明为：

torch.Tensor.view(*shape) → Tensor

函数作用为：

Returns a new tensor with the same data as the self tensor but of a different shape.

即为保持原张量的数据不变，变化张量的形状。

>>> x = torch.randn(4, 4)
>>> x.size()
torch.Size([4, 4])
>>> y = x.view(16)
>>> y.size()
torch.Size([16])
>>> z = x.view(-1, 8)  # the size -1 is inferred from other dimensions
>>> z.size()
torch.Size([2, 8])

torch.index_select

函数声明为：

torch.index_select(input, dim, index, out=None) → Tensor

函数的作用为：

Returns a new tensor which indexes the input tensor along dimension dim using the entries in index which is a LongTensor.

也就是返回一个新的张量，相对于原张量来说，这个张量的其他维度保持不变，但是其选定的维度dim为仅为通过index张量所选定的那些维。index张量是一个一维的向量。

举个例子来说：

>>> x = torch.randn(3, 4)
>>> x
tensor([[ 0.1427,  0.0231, -0.5414, -1.0009],
        [-0.4664,  0.2647, -0.1228, -1.1068],
        [-1.1734, -0.6571,  0.7230, -0.6004]])
>>> indices = torch.tensor([0, 2])
>>> torch.index_select(x, 0, indices)
tensor([[ 0.1427,  0.0231, -0.5414, -1.0009],
        [-1.1734, -0.6571,  0.7230, -0.6004]])
>>> torch.index_select(x, 1, indices)
tensor([[ 0.1427, -0.5414],
        [-0.4664, -0.1228],
        [-1.1734,  0.7230]])

注意，这个函数与torch.gather函数的区别：index_select函数中index参数的值仅仅代表了张量在这一维度上所有张量的index，而gather函数中的index参数的值代表这一维度上所有元素的index，举一个二维的例子：

torch.max

函数声明为：

torch.max(input, dim, keepdim=False, out=None) -> (Tensor, LongTensor)

函数作用为：

Returns a namedtuple (values, indices) where values is the maximum value of each row of the input tensor in the given dimension dim. And indices is the index location of each maximum value found (argmax).

举个例子来说：

>>> a = torch.randn(4, 4)
>>> a
tensor([[-1.2360, -0.2942, -0.1222,  0.8475],
        [ 1.1949, -1.1127, -2.2379, -0.6702],
        [ 1.5717, -0.9207,  0.1297, -1.8768],
        [-0.6172,  1.0036, -0.6060, -0.2432]])
>>> torch.max(a, 1)
torch.return_types.max(values=tensor([0.8475, 1.1949, 1.5717, 1.0036]), indices=tensor([3, 0, 0, 1]))

torch.transpose

函数声明为：

torch.transpose(input, dim0, dim1) → Tensor

函数作用为：

Returns a tensor that is a transposed version of input. The given dimensions dim0 and dim1 are swapped.

举个例子来说：

>>> x = torch.randn(2, 3)
>>> x
tensor([[ 1.0028, -0.9893,  0.5809],
        [-0.1669,  0.7299,  0.4942]])
>>> torch.transpose(x, 0, 1)
tensor([[ 1.0028, -0.1669],
        [-0.9893,  0.7299],
        [ 0.5809,  0.4942]])

torch.cat

函数声明为：

torch.cat(tensors, dim=0, out=None) → Tensor

函数作用为：

Concatenates the given sequence of seq tensors in the given dimension. All tensors must either have the same shape (except in the concatenating dimension) or be empty.

举个例子来说：

>>> x = torch.randn(2, 3)
>>> x
tensor([[ 0.6580, -1.0969, -0.4614],
        [-0.1034, -0.5790,  0.1497]])
>>> torch.cat((x, x, x), 0)
tensor([[ 0.6580, -1.0969, -0.4614],
        [-0.1034, -0.5790,  0.1497],
        [ 0.6580, -1.0969, -0.4614],
        [-0.1034, -0.5790,  0.1497],
        [ 0.6580, -1.0969, -0.4614],
        [-0.1034, -0.5790,  0.1497]])
>>> torch.cat((x, x, x), 1)
tensor([[ 0.6580, -1.0969, -0.4614,  0.6580, -1.0969, -0.4614,  0.6580,
         -1.0969, -0.4614],
        [-0.1034, -0.5790,  0.1497, -0.1034, -0.5790,  0.1497, -0.1034,
         -0.5790,  0.1497]])