PyTorch笔记1----------Tensor(张量):基本概念、创建、属性、算数运算

1.Tensor的基本概念

在PyTorch中，张量可以看作一个多维数组。标量是零维的张量，向量是一维的张量，矩阵是二维的张量。

2.Tensor的创建

• 直接初始化值

import torch
a = torch.Tensor([[1,2],[3,4]])
print(a)

• 定义tensor的形状创建，此时值未初始化，是随机值

import torch
a = torch.Tensor(2,3)
print(a)

• 值全为1 的tensor

import torch
a = torch.ones(2,3)
print(a)

• 对角线值全为1的tensor，其他位置值为0

import torch
a = torch.eye(2,2)
print(a)

• 值全为0的tensor

import torch
a = torch.zeros(2,2)
print(a)

• 一个和现有tensor形状相同的值全为0的tensor

import torch
a = torch.Tensor(2,3)
b = torch.zeros_like(a)
print('a:\n',a)
print('b:\n',b)

• 一个和现有tensor形状相同的值全为1的tensor

import torch
a = torch.Tensor(2,3)
b = torch.ones_like(a)
print('a:\n',a)
print('b:\n',b)

• 随机值（0～1）

import torch
a = torch.rand(2,3)
print(a)

• 满足正态分布

import torch
#mean:表示均值，std:表示标准差
a = torch.normal(mean=torch.rand(5),std=torch.rand(5))
print(a)

• 满足均匀分布

import torch
# uniform_: 从一个均匀分布-1到1之间随机采样
a = torch.Tensor(2,2).uniform_(-1,1)
print(a)

• 等差序列

import torch
a = torch.arange(0,10,2)
print(a)

• 等间隔的值

import torch
a = torch.linspace(2,10,3)
print(a)

• 乱序

import torch
a = torch.randperm(10)
print(a)

3.Tensor的属性

• 将张量定义在CPU或GPU

import torch
dev1 = torch.device("cpu")
dev2 = torch.device("cuda:0")
a = torch.tensor([2,2],device=dev1)
b = torch.tensor([2,2],device=dev2)
print("a:",a)
print("b:",b)

• 定义类型

import torch
a = torch.tensor([2,2],dtype=torch.float32)
print("a:",a)

• 稀疏的张量：torch.sparse_coo_tensor

import torch
#坐标
i = torch.tensor([[0,1,2],[0,1,2]])
v = torch.tensor([1,2,3])
a = torch.sparse_coo_tensor(i,v,(4,4))
print("a:",a)

• 稀疏张量转稠密张量

import torch
#坐标
i = torch.tensor([[0,1,2],[0,1,2]])
v = torch.tensor([1,2,3])
a = torch.sparse_coo_tensor(i,v,(4,4)).to_dense()
print("a:",a)

4.Tensor的算数运算

• 加法运算

import torch
a = torch.rand(2,3)  
b = torch.rand(2,3)
print('初始a:\n',a)
print('b:\n',b)
print("a+b:\n",a + b)
print("a.add(b):\n",a.add(b))
print("torch.add(a,b):\n",torch.add(a,b))
print("a.add_(b):\n",a.add_(b))
print("经过a.add_(b)后的a:\n",a)

• 减法运算

import torch
a = torch.rand(2,3)  
b = torch.rand(2,3)
print('初始a:\n',a)
print('b:\n',b)
print("a-b:\n",a - b)
print("a.sub(b):\n",a.sub(b))
print("torch.sub(a,b):\n",torch.sub(a,b))
print("a.sub_(b):\n",a.sub_(b))
print("经过a.sub_(b)后的a:\n",a)

• 乘法运算

import torch
a = torch.rand(2,3)  
b = torch.rand(2,3)
print('初始a:\n',a)
print('b:\n',b)
print("a*b:\n",a * b)
print("a.mul(b):\n",a.mul(b))
print("torch.mul(a,b):\n",torch.mul(a,b))
print("a.mul_(b):\n",a.mul_(b))
print("经过a.mul_(b)后的a:\n",a)

• 除法运算

import torch
a = torch.rand(2,3)  
b = torch.rand(2,3)
print('初始a:\n',a)
print('b:\n',b)
print("a/b:\n",a / b)
print("a.div(b):\n",a.div(b))
print("torch.div(a,b):\n",torch.div(a,b))
print("a.div_(b):\n",a.div_(b))
print("经过a.div_(b)后的a:\n",a)

• 矩阵运算

import torch
a = torch.ones(2,1)  
b = torch.ones(1,2)
print('初始a:\n',a)
print('b:\n',b)
print("a@b:\n",a @ b)
print("a.matmul(b):\n",a.matmul(b))
print("torch.matmul(a,b):\n",torch.matmul(a,b))
print("torch.mm(a,b):\n",torch.mm(a,b))
print("a.mm(b):\n",a.mm(b))

• 高维tensor(两组数据必须保证最后两个维度可以进行矩阵运算)

import torch
a = torch.ones(1,2,3,4)  
b = torch.ones(1,2,4,3)
print('初始a:\n',a)
print('b:\n',b)
print("a@b:\n",a @ b)
print("a.matmul(b):\n",a.matmul(b))
print("torch.matmul(a,b):\n",torch.matmul(a,b))

• 指数运算

import torch
a = torch.tensor([1,2])
print('初始a:\n',a)
print("torch.pow(a,3):\n",torch.pow(a,3))
print("a.pow(3):\n",a.pow(3))
print("a**3:\n",a**3)
print("a.pow_(3):\n",a.pow_(3))
print("经过a.pow_(3)后的a:\n",a)

• 关于自然函数e的相关运算

import torch
a = torch.tensor([1,2],dtype=torch.float32)
print('初始a:\n',a)
print("torch.exp(a):\n",torch.exp(a))
print("torch.exp_(a):\n",torch.exp_(a))
print("经过torch.exp_(a)后的a:\n",a)
print("a.exp():\n",a.exp())
print("a.exp_():\n",a.exp_())
print("经过a.exp_()后的a:\n",a)

• 对数运算

import torch
a = torch.tensor([10,2],dtype=torch.float32)
print('初始a:\n',a)
#以e为底
print("torch.log(a):\n",torch.log(a))
print("torch.log_(a):\n",torch.log_(a))
print("经过torch.log_(a)后的a:\n",a)
print("a.log():\n",a.log())
print("a.log_():\n",a.log_())
print("经过a.log_()后的a:\n",a)

• 平方根运算

import torch
a = torch.tensor([10,2],dtype=torch.float32)
print('初始a:\n',a)
#以e为底
print("torch.sqrt(a):\n",torch.sqrt(a))
print("torch.sqrt_(a):\n",torch.sqrt_(a))
print("经过torch.sqrt_(a)后的a:\n",a)
print("a.sqrt():\n",a.sqrt())
print("a.sqrt_():\n",a.sqrt_())
print("经过a.sqrt_()后的a:\n",a)

知识点为听课总结笔记，课程为B站"2025最新整合！公认B站讲解最强【PyTorch】入门到进阶教程，从环境配置到算法原理再到代码实战逐一解读，比自学效果强得多！"：2025最新整合！公认B站讲解最强【PyTorch】入门到进阶教程，从环境配置到算法原理再到代码实战逐一解读，比自学效果强得多！_哔哩哔哩_bilibili