torch.Tensor和numpy.ndarray

1. torch.Tensor和numpy.ndarray相互转换

import torch

import numpy as np

# <class 'numpy.ndarray'>

np_data = np.arange(6).reshape((2,3))

# <class 'torch.Tensor'>

torch_data = torch.from_numpy(np_data)

# <class 'numpy.ndarray'>

tensor2array = torch_data.numpy()

print('numpy array:\n',np_data,type(np_data),

      '\ntorch tensor:\n',torch_data,type(torch_data),

      '\ntensor to array:\n',tensor2array,type(tensor2array))

# numpy array:
#  [[0 1 2]
#  [3 4 5]] <class 'numpy.ndarray'>
# torch tensor:
#  tensor([[0, 1, 2],
#         [3, 4, 5]]) <class 'torch.Tensor'>
# tensor to array:
#  [[0 1 2]
#  [3 4 5]] <class 'numpy.ndarray'>

torch.Tensor:是一个包含了一种数据类型元素的多维矩阵，缺省为torch.FloatTensor
2. torch.Tensor和numpy.ndarray一些简单操作，如均值，绝对值，sin，log等

data = [-1,-2,1,2]

tensor_default = torch.Tensor(data)

tensor = torch.FloatTensor(data)

print('tensor default type:\n',tensor_default,

      '\ntensor FloatTensor type:\n',tensor,

      '\nabs:',

      '\nnumpy:',np.abs(data),

      '\ntorch:',torch.abs(tensor),

      '\nsin:',

      '\nnumpy:',np.sin(data),

      '\ntorch:',torch.sin(tensor),

      '\nmean:',

      '\nnumpy:',np.mean(data),

      '\ntorch:',torch.mean(tensor),)

# tensor default type:
#  tensor([-1., -2.,  1.,  2.])
# tensor FloatTensor type:
#  tensor([-1., -2.,  1.,  2.])
# abs:
# numpy: [1 2 1 2]
# torch: tensor([1., 2., 1., 2.])
# sin:
# numpy: [-0.84147098 -0.90929743  0.84147098  0.90929743]
# torch: tensor([-0.8415, -0.9093,  0.8415,  0.9093])
# mean:
# numpy: 0.0
# torch: tensor(0.)

3. 矩阵乘法(正确的做法)

data = [[1,2], [3,4]]

tensor = torch.FloatTensor(data)

print(

    '\nmatrix multiplication (matmul):',

    '\nnumpy:\n', np.matmul(data, data),     # [[7, 10], [15, 22]]

    '\ntorch:\n', torch.mm(tensor, tensor))  # [[7, 10], [15, 22]]

# matrix multiplication (matmul):
# numpy:
#  [[ 7 10]
#  [15 22]]
# torch:
#  tensor([[ 7., 10.],
#         [15., 22.]])