pytorch基础学习1

1.从numpy导入

import torch
import numpy as np
a=np.array([2,2.3])
b=torch.from_numpy(a)
print(a)
print(b)
a=np.ones([2,3])
b=torch.from_numpy(a)
print(a)
print(b)

结果：[2.  2.3]
tensor([2.0000, 2.3000], dtype=torch.float64)
[[1. 1. 1.]
 [1. 1. 1.]]
tensor([[1., 1., 1.],
        [1., 1., 1.]], dtype=torch.float64)

2.tensor一般具体向量，FloatTensor一般是维度，也可以向量

import torch
b=torch.tensor([2,2.3])
print(b)
b=torch.FloatTensor(2,3)
print(b)

结果：

tensor([2.0000, 2.3000])
tensor([[2.1998e-31, 0.0000e+00, 1.4013e-45],
        [0.0000e+00, 1.4013e-45, 0.0000e+00]])

3.设置默认值

torch.set_default_tensor_type(torch.DoubleTensor)

4.rand根据size创建[0,1]的随机数tensor,randint（min，max，size...

import torch
b=torch.rand(3,3)
print(b)
b=torch.randint(0,10,(5,))
print(b)

结果：tensor([[0.0576, 0.6674, 0.2087],
        [0.7914, 0.3789, 0.0539],
        [0.6894, 0.6893, 0.3720]])
tensor([2, 3, 6, 9, 2])

5.randn根据size创建标准正太分布的随机数tensor，均值0，方差1

import torch
b=torch.randn(3,3)
print(b)

结果：tensor([[-0.4682, -0.2454,  1.6473],
        [ 0.7681,  1.7327,  0.0364],
        [-0.6486, -0.2380,  0.2747]])

6.full

import torch
b=torch.full([2,3],3)
print(b)
b=torch.full([],3)
print(b)
b=torch.full([1],3)
print(b)
b=torch.full([2],3)
print(b)

结果：

tensor([[3., 3., 3.],
        [3., 3., 3.]])
tensor(3.)
tensor([3.])
tensor([3., 3.])

7.linspace（start，end，step）

import torch
b=torch.linspace(0,10,3)
print(b)

结果：tensor([ 0.,  5., 10.])

8.arange

import torch
b=torch.arange(0,10)
print(b)
b=torch.arange(0,10,3)
print(b)

结果：tensor([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
tensor([0, 3, 6, 9])

9.torch.ones，torch.zeros，torch.eye

import torch
b=torch.ones(3,3)
print(b)
b=torch.zeros(3,3)
print(b)
b=torch.eye(3,4)
print(b)

结果：tensor([[1., 1., 1.],
        [1., 1., 1.],
        [1., 1., 1.]])
tensor([[0., 0., 0.],
        [0., 0., 0.],
        [0., 0., 0.]])
tensor([[1., 0., 0., 0.],
        [0., 1., 0., 0.],
        [0., 0., 1., 0.]])