基于pytorch的线性回归模型，python

import torch from torch import nn from torch.utils.data import TensorDataset, DataLoader import pandas as pd import numpy as np import matplotlib.pyplot as plt #读取含有样本数据的CSV文件 df = pd.read_csv('example.csv') new_index = np.random.permutation(df.index) # 打乱数据集 df = df.reindex(new_index).reset_index(drop=True) # 设置训练集比例 train_dataset_rate=0.8 num_train = int(len(df)*train_dataset_rate) #数据集最后一列是因变量 X_train = df.iloc[:num_train, :-1] # 训练集特征 y_train = df.iloc[:num_train, -1] # 训练集标注结果 X_test = df.iloc[num_train:, :-1] # 测试集特征 y_test = df.iloc[num_train:, -1] # 测试集标注结果 X_train_tensor = torch.tensor(X_train.values, dtype=torch.float32) y_train_tensor = torch.tensor(y_train.values, dtype=torch.float32) y_train_tensor = y_train_tensor.reshape((-1,1)) X_test_tensor = torch.tensor(X_test.values, dtype=torch.float32) y_test_tensor = torch.tensor(y_test.values, dtype=torch.float32) y_test_tensor = y_test_tensor.reshape((-1,1)) # 使用TensorDataset,DataLoader创建训练集 dataset_train = TensorDataset(X_train_tensor, y_train_tensor) loader_train = DataLoader(dataset_train, batch_size=4, shuffle=True) # 迭代次数 epochs=200 # 构建线性回归网络 net = nn.Sequential(nn.Linear(X_train_tensor.shape[1],1)) # 损失函数 loss = nn.MSELoss() #优化器 optimizer = torch.optim.SGD(net.parameters(),lr=0.3) # 初始化一个列表来存储每次迭代的loss值 loss_list = [] # 训练 for epoch in range(epochs): for _,(X,y) in enumerate(loader_train): optimizer.zero_grad() l = loss(net(X),y) l.backward() optimizer.step() loss_list.append(loss(net(X_train_tensor),y_train_tensor).data) print(f'第{epoch+1}次的loss值为{loss(net(X_train_tensor),y_train_tensor).data}') print("训练完成") print("每个特征值的系数分别为") print(net[0].weight.data[0].numpy()) print("偏置b为") print(net[0].bias.data[0].numpy()) # loss曲线 plt.plot(loss_list) plt.title('Loss Over Epochs') plt.xlabel('Epoch') plt.ylabel('Loss') plt.grid(True) plt.show() # 对测试集的预测结果 y_pred=net(X_test_tensor) x = np.arange(len(y_pred)) plt.plot(x,y_test_tensor.numpy(),label='Actual Values') plt.scatter(x,y_pred.data.numpy(),c='red',label='Predicted Values') plt.legend() plt.title('Comparison of Actual and Predicted Values') plt.xlabel('Index') plt.ylabel('Value') plt.grid(True) plt.show()