pytorch 常用函数

pytorch 常用函数

参考：网易云课堂pytorch 学习

#创建tensor import from numpy
import numpy as np
import torch
a=np.array([2,3.3])
torch.from_numpy(a) # out:tensor([2.0000, 3.3000], dtype=torch.float64)

a=np.ones([2,3])
torch.from_numpy(a) 
'''tensor([[1., 1., 1.],
        [1., 1., 1.]], dtype=torch.float64)''' 


#import from List

torch.tensor([2,3.0])#out:tensor([2., 3.]) tensor 加载的数据

torch.FloatTensor([2,3.2])# tensor([2.0000, 3.2000])

torch.tensor([[2.,3.2],[1.0,22.3]])  #tensor([[ 2.0000,  3.2000],
                                      #[ 1.0000, 22.3000]])


#uninitialized
#1、Torch.empty()
#2、Torch.FloatTensor(d1, d2, d3)
#3、 Torch.IntTensr(d1, d2, d3)

torch.empty(1)#输出为1*1 未初始化的矩阵 out:tensor([0.])

torch.Tensor(2,3)

'''
tensor([[0., 0., 0.],
        [0., 0., 0.]])'''

torch.IntTensor(2,3)
'''tensor([[0, 0, 0],
        [0, 0, 0]], dtype=torch.int32)'''

torch.FloatTensor(2,3)

'''tensor([[0., 0., 0.],
        [0., 0., 0.]])'''


#set default type

torch.tensor([1.2,3]).type() # pytorch 默认tensor 'torch.FloatTensor' 

torch.set_default_tensor_type(torch.DoubleTensor)

torch.tensor([1.2,3]).type()# type()是方法  'torch.DoubleTensor'

#rand /rand_like,randint
torch.rand(3,3)
'''tensor([[0.9379, 0.9233, 0.1059],
        [0.9286, 0.3901, 0.3092],
        [0.5107, 0.7357, 0.7802]])'''


a=torch.rand(3,3)
torch.rand_like(a)

torch.randint(1,10,[2,3]) #min,max,size
# tensor([[8, 5, 8],
#         [1, 9, 6]])


#正态分布
#N(0, 1)
#N(u, std) 


torch.randn(3,3)#正态分布

'''tensor([[8, 5, 8],
        [1, 9, 6]])'''

torch.normal(mean=torch.full([10],0),std=torch.arange(1,0,-0.1))
'''tensor([-0.6110, -0.6861, -0.8501,  0.4197, -0.2862,  0.0783, -0.4763,  0.2206,
        -0.1437, -0.0090]) 均值，方差 torch.arange 左闭右开区间'''
torch.normal(mean=torch.full([10],0),std=torch.arange(1,0,-0.1))
'''tensor([ 1.6395, -1.2369,  0.9526, -0.1463,  0.1491, -0.1538,  0.3134, -0.2822,
        -0.0860, -0.0185])'''



#full

torch.full([2,3],7)
'''tensor([[7., 7., 7.],
        [7., 7., 7.]]) 2维'''

torch.full([],7) #scalar标量  tensor(7.)

torch.full([1],7)# out:tensor([7.])


#arange /range
torch.arange(0,10)# tensor([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

torch.arange(0,10,2) #tensor([0, 2, 4, 6, 8]) 左闭右开 步长

torch.range(0,10) # 左闭右开区间tensor([ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.,  9., 10.]) 


#linspace/logspace

torch.linspace(0,10,steps=4)#左闭右开数量tensor([ 0.0000,  3.3333,  6.6667, 10.0000])

torch.linspace(0,10,steps=10) #tensor([ 0.0000,  1.1111,  2.2222,  3.3333,  4.4444,  5.5556,  6.6667,  7.7778,8.8889, 10.0000])

torch.logspace(0,-1,steps=10)#logspace以 2，10，e 为底的指数tensor([1.0000, 0.7743, 0.5995, 0.4642, 0.3594, 0.2783, 0.2154, 0.1668, 0.1292,0.1000])

torch.logspace(0,1,steps=10) #tensor([ 1.0000,  1.2915,  1.6681,  2.1544,  2.7826,  3.5938,  4.6416,  5.9948,7.7426, 10.0000])


#Ones/zeros/eye 
torch.ones(3,3) #3*3的全一矩阵
'''
tensor([[1., 1., 1.],
        [1., 1., 1.],
        [1., 1., 1.]])'''

torch.zeros(3,3)#3*3 的全0矩阵
'''
tensor([[0., 0., 0.],
        [0., 0., 0.],
        [0., 0., 0.]])'''
torch.eye(3,4) #单位矩阵
'''tensor([[1., 0., 0., 0.],
        [0., 1., 0., 0.],
        [0., 0., 1., 0.]])'''
torch.eye(3)
'''tensor([[1., 0., 0.],
        [0., 1., 0.],
        [0., 0., 1.]])'''
a=torch.zeros(3,3)
torch.ones_like(a)
'''tensor([[1., 1., 1.],
        [1., 1., 1.],
        [1., 1., 1.]])'''


#randperm random.shuffle 对索引随机打散
a=torch.rand(2,3) #类似第一维为人名  a 人名 数学成绩 ，b 人名 语文成绩 两个矩阵的行的数目一样 随机种子数目为行的数目 [0,1]为原始顺序，[1,0]为反转顺序
b=torch.rand(2,2)
idx=torch.randperm(2)
idx1=idx#tensor([0, 1])
print(idx1)
idx2=idx1#tensor([1, 0])
print(idx2)
a #'''tensor([[0.6799, 0.1168, 0.6706], [0.2596, 0.6144, 0.6370]])'''
print(a[[0,1]]) #tensor([[0.1919, 0.0593, 0.7492],
        #[0.7986, 0.2391, 0.8857]])
print(a[[1,0]])
print(b[idx]) #tensor([[0.2905, 0.9161],
        #[0.6948, 0.3815]])
a,b
'''
tensor([1, 0])
tensor([1, 0])
tensor([[0.4416, 0.2493, 0.6051],
        [0.2301, 0.5072, 0.3799]])
tensor([[0.2301, 0.5072, 0.3799],
        [0.4416, 0.2493, 0.6051]])
tensor([[0.4023, 0.0007],
        [0.3239, 0.4355]])
(tensor([[0.4416, 0.2493, 0.6051],
         [0.2301, 0.5072, 0.3799]]), tensor([[0.3239, 0.4355],
         [0.4023, 0.0007]]))'''