Pytorch 实现线性回归

Pytorch 实现线性回归

import torch
from torch.utils import data
from torch import nn

# 合成数据
def synthetic_data(w, b, num_examples):
    """y = Xw + b + zs"""
    X = torch.normal(0, 1, (num_examples, len(w)))
    y = torch.matmul(X, w) + b
    y += torch.normal(0, 0.01, y.shape)
    return X, y.reshape((-1, 1))

# 用于合成数据的模板
true_w = torch.tensor([2, -3.4, 2])
true_b = 4.2

# 合成1000个数据
features, labels = synthetic_data(true_w, true_b, 1000)

# 随机批量加载数据
def load_array(data_arrays, batch_size, is_train=True):
    dataset = data.TensorDataset(*data_arrays)
    return data.DataLoader(dataset, batch_size, shuffle=is_train)

batch_size = 10
data_iter = load_array((features, labels), batch_size)

# 初始化线性网络，3输入1输出
net = nn.Sequential(nn.Linear(3, 1))
# 均方误差损失函数
loss = nn.MSELoss()
# 优化算法
trainer = torch.optim.SGD(net.parameters(), lr=0.03)

# 开始迭代
num_epochs = 3
for epoch in range(num_epochs):
    for X, y in data_iter:
        l = loss(net(X), y)
        trainer.zero_grad()
        l.backward()
        trainer.step()
    l = loss(net(features), labels)
    print(f'epoch {epoch + 1}, loss {l:f}')