pytorch学习03：张量数据类型和一些操作

张量数据类型

Data type	dtype	CPU tensor	GPU tensor
32-bit floating point	torch.float32 or torch.float	torch.FloatTensor	torch.cuda.FloatTensor
64-bit floating point	torch.float64 or torch.double	torch.DoubleTensor	torch.cuda.DoubleTensor
16-bit floating point	torch.float16 or torch.half	torch.HalfTensor	torch.cuda.HalfTensor
8-bit integer(unsigned)	torch.uint8	torch.ByteTensor	torch.cuda.ByteTensor
8-bit integer(signed)	torch.int8	torch.CharTensor	torch.cuda.CharTensor
16-bit integer(signed)	torch.int16 or torch.short	torch.ShortTensor	torch.cuda.ShortTensor
32-bit integer(signed)	torch.int32 or torch.int	torch.IntTensor	torch.cuda.IntTensor
64-bit integer(signed)	torch.int64 or torch.long	torch.LongTensor	torch.cuda.LongTensor

类型使用

示例1：类型比较

import torch

# 随机初始化
a = torch.randn(2, 3)

print("a.type(): ", a.type())
print("type(a): ", type(a))
print("type of 'a' is torch.FloatTensor: ",
      isinstance(a, torch.FloatTensor))

示例2：同一数据被部署在CPU和GPU上类型不同

import torch

# 随机初始化
a = torch.randn(2, 3)

print(isinstance(a, torch.cuda.FloatTensor))
# 如果没有开启cuda会报错
a = a.cuda()
print(isinstance(a, torch.cuda.FloatTensor))

示例3：标量

import torch

a = torch.tensor(1.)
print("a: ", a)

b = torch.tensor(1.3)
print("b: ", b)

示例4：标量的shape

import torch

a = torch.tensor(2.2)
print("a.shape: ", a.shape)
print("len(a.shape): ", len(a.shape))
print("a.dim(): ", a.dim())
print("a.size(): ", a.size())

示例5：一维张量

import numpy as np
import torch

print(torch.tensor([1.1]))
print(torch.tensor([1.1, 2.2]))
print(torch.FloatTensor(1))
print(torch.FloatTensor(2))

data = np.ones(2)
print("data: ", data)
data = torch.from_numpy(data)
print("tensor from numpy:", data)

print("data.shape: ", data.shape)
print("data.dim(): ", data.dim())
print("data.size(): ", data.size())

示例6：二位张量

import torch

a = torch.randn(2, 3)
print("a:", a)
print("a.shape:", a.shape)
print("a.size():", a.size())
print("a.size(0):", a.size(0))
print("a.size(1):", a.size(1))
print("a.shape[1]:", a.shape[1])
print("a.dim():", a.dim())

示例7：三维张量

import torch

# 随机均匀分布初始化
a = torch.rand(2, 2, 3)

print("a:", a)
print("list(a.shape):", list(a.shape))
print("a.size():", a.size())
print("a.size(0):", a.size(0))
print("a.size(1):", a.size(1))
print("a.shape[1]:", a.shape[1])
print("a.dim():", a.dim())

示例8：获得总元素个数

import torch

# 随机均匀分布初始化
a = torch.rand(2, 2, 3)

# 返回a的总元素个数 2 * 2 * 3
print("a.numel():", a.numel())