pytorch包学习--touch.Tensor

文章来自于翻译https://pytorch.org/docs/stable/tensors.html，想自己学习，可以参考原文

1.torch.tensor 是一个包含一种数据的多维矩阵。

2.new_full(size, fill_value, dtype=None, device=None, requires_grad=False) → Tensor

Returns a Tensor of size size filled with fill_value. By default, the returned Tensor has the same torch.dtype and torch.device as this tensor.

3.new_tensor(data, dtype=None, device=None, requires_grad=False) → Tensor

Returns a new Tensor with data as the tensor data. By default, the returned Tensor has the same torch.dtype and torch.device as this tensor.

new_tensor() always copies data. If you have a Tensor data and want to avoid a copy, use torch.Tensor.requires_grad_() or torch.Tensor.detach(). If you have a numpy array and want to avoid a copy, use torch.from_numpy().

4.reshape(*shape) → Tensor

Returns a tensor with the same data and number of elements as self but with the specified shape. This method returns a view if shape is compatible with the current shape. See torch.Tensor.view() on when it is possible to return a view.

5.reshape_as(other) → Tensor

Returns this tensor as the same shape as other. self.reshape_as(other) is equivalent to self.reshape(other.sizes()). This method returns a view if other.sizes() is compatible with the current shape. See torch.Tensor.view() on when it is possible to return a view.

6.scatter_add_(dim, index, other) → Tensor

Adds all values from the tensor other into self at the indices specified in the indextensor in a similar fashion as scatter_(). For each value in other, it is added to an index in self which is specified by its index in other for dimension != dim and by the corresponding value in index for dimension = dim.

For a 3-D tensor, self is updated as:

self[index[i][j][k]][j][k] += other[i][j][k]  # if dim == 0
self[i][index[i][j][k]][k] += other[i][j][k]  # if dim == 1
self[i][j][index[i][j][k]] += other[i][j][k]  # if dim == 2

示例如下:
>>> x = torch.rand(2, 5)
>>> x
tensor([[0.7404, 0.0427, 0.6480, 0.3806, 0.8328],
        [0.7953, 0.2009, 0.9154, 0.6782, 0.9620]])
>>> torch.ones(3, 5).scatter_add_(0, torch.tensor([[0, 1, 2, 0, 0], [2, 0, 0, 1, 2]]), x)
tensor([[1.7404, 1.2009, 1.9154, 1.3806, 1.8328],
        [1.0000, 1.0427, 1.0000, 1.6782, 1.0000],
        [1.7953, 1.0000, 1.6480, 1.0000, 1.9620]])

7.to(*args, **kwargs) → Tensor

Performs Tensor dtype and/or device conversion

转数据类型,和cpu,gpu切换.

>>> tensor = torch.randn(2, 2)  # Initially dtype=float32, device=cpu
>>> tensor.to(torch.float64)
tensor([[-0.5044,  0.0005],
        [ 0.3310, -0.0584]], dtype=torch.float64)

>>> cuda0 = torch.device('cuda:0')
>>> tensor.to(cuda0)
tensor([[-0.5044,  0.0005],
        [ 0.3310, -0.0584]], device='cuda:0')

>>> tensor.to(cuda0, dtype=torch.float64)
tensor([[-0.5044,  0.0005],
        [ 0.3310, -0.0584]], dtype=torch.float64, device='cuda:0')

>>> other = torch.randn((), dtype=torch.float64, device=cuda0)
>>> tensor.to(other, non_blocking=True)
tensor([[-0.5044,  0.0005],
        [ 0.3310, -0.0584]], dtype=torch.float64, device='cuda:0')

8.view(*shape) → Tensor

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

The returned tensor shares the same data and must have the same number of elements, but may have a different size. For a tensor to be viewed, the new view size must be compatible with its original size and stride

改变底层数据

>>> 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])

9.all() → bool

Returns True if all elements in the tensor are non-zero, False otherwise.