手动实现线性回归

import random
import torch
from d2l import torch as d2l
import matplotlib.pyplot as plt

def synthetic_data(w,b,num_example):
    #生成带噪声的数据
    x = torch.normal(0,1,(num_example,len(w)));
    y = torch.matmul(x,w) +b
    y = y + torch.normal(0,0.01,y.shape)
    return x,y.reshape((-1,1))

true_w = torch.tensor([2,-3.4])
true_b = 4.2
features, labels = synthetic_data(true_w,true_b,1000)

#1.展示数据
# plt.figure()
# plt.scatter(features[:,(1)].detach().numpy(),labels.detach().numpy(),1)  #将torch转换成numpy
# plt.show()

# print(type(labels),type(labels.detach()))   <class 'torch.Tensor'> <class 'torch.Tensor'>

#2.读取数据
#我们需要对数据集进行遍历，每次取一次小批量的样本，使用它们来更新我们的模型
# 定义data_iter   输入参数:批量大小、特征矩阵、标签向量 生成batch_size的小批量
#每个小批量包含一组特征和标签

def data_iter(batch_size,features,labels):
    num_examples = len(features)
    indices = list(range(num_examples))
    random.shuffle(indices)

    for i in range(0,num_examples,batch_size):
        batch_indices = torch.tensor(
            indices[i:min(i+batch_size,num_examples)]
        )
        yield features[batch_indices],labels[batch_indices]   #类似于return，只不过返回的是生成器
        #用return就会报错

batch_size = 10
#执行效率低 要求数据必须在内存中，并执行大量的内存随机访问
# for x,y in data_iter(batch_size,features,labels):
#     print(x,'\n',y)
#     break;

#初始化模型参数
# w = torch.zeros(size=(2,1), requires_grad=True)    #参数为0的时候也可以
w = torch.normal(0, 0.01, size=(2,1), requires_grad=True)
b = torch.zeros(1, requires_grad=True)

#定义模型
def linreg(x,w,b):
    return torch.matmul(x,w)+b

#定义损失函数
def squared_loss(y_hat,y):
    return (y_hat - y.reshape(y_hat.shape))**2/2

#定义优化算法
def sgd(params , lr ,batch_size):
    with torch.no_grad():
        for param in params:
            param -=lr*param.grad / batch_size
            param.grad.zero_()


#训练
lr = 0.03
num_epochs = 3
net = linreg
loss = squared_loss

for epoch in range(num_epochs):
    for x,y in data_iter(batch_size,features,labels):
        l = loss(net(x,w,b),y)
        # 因为l形状是(batch_size,1)，而不是一个标量
        l.sum().backward()
        sgd([w, b], lr, batch_size)
    with torch.no_grad(): #强调不进行计算图生成
        train_l = loss(net(features, w, b), labels)
        print(f'epoch {epoch + 1}, loss {float(train_l.mean()):f}')

print(f'w的估计误差: {true_w - w.reshape(true_w.shape)}')
print(f'b的估计误差: {true_b - b}')