《PyTorch深度学习实践》08 加载数据集

1. 说明

本系列博客记录B站课程《PyTorch深度学习实践》的实践代码课程链接请点我

2. 代码

# ---------------------------
# @Time     : 2022/4/14 11:45
# @Author   : lcq
# @File     : 08_MyDataLoaderTest.py
# @Function : 
# ---------------------------

import torch
import numpy as np
from torch.utils.data import Dataset
from torch.utils.data import DataLoader
from sklearn.model_selection import train_test_split

# Step1: 加载数据
filePath = "data//diabetes.csv"
xy = np.loadtxt(filePath, delimiter=',', dtype=np.float32)
X = xy[:, :-1]
Y = xy[:, [-1]]
X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.3)    # 划分测试集和训练集，test_size=0.3为测试集的比例
X_test = torch.from_numpy(X_test)       # 转换程torch的张量
Y_test = torch.from_numpy(Y_test)


# Step2：将训练集进行小批量处理
class MiniBatchDataset(Dataset):
    def __init__(self, X_input, Y_label):
        self.len = X_input.shape[0]
        self.X_input = torch.from_numpy(X_input)
        self.Y_label = torch.from_numpy(Y_label)

    def __getitem__(self, index):
        return self.X_input[index], self.Y_label[index]

    def __len__(self):
        return self.len


# Step3: 将数据变成小批量数据
train_data = MiniBatchDataset(X_train, Y_train)
train_loader = DataLoader(dataset=train_data, batch_size=32, shuffle=True, num_workers=2)


# Step4: 构建模型
class Model(torch.nn.Module):
    def __init__(self, column):
        super(Model, self).__init__()
        self.model = torch.nn.Sequential(
            torch.nn.Linear(column, 6),
            torch.nn.Sigmoid(),
            torch.nn.Linear(6, 4),
            torch.nn.Sigmoid(),
            torch.nn.Linear(4, 2),
            torch.nn.Sigmoid(),
            torch.nn.Linear(2, 1),
            torch.nn.Sigmoid()
        )

    def forward(self, X_input):
        y_pred = self.model(X_input)
        return y_pred


model = Model(column=X_test.shape[1])


# Step5: 构建损失函数和优化器
Loss = torch.nn.BCELoss(reduction='mean')
optimizer = torch.optim.SGD(model.parameters(), lr=0.1)


# Step6: 训练
def train(epoch):
    for i, data in enumerate(train_loader, 0):
        inputs, labels = data
        y_pred = model.forward(inputs)
        loss = Loss(y_pred, labels)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        print("epoch: ", epoch, "i: ", i, "loss: ", loss.item())


def test():
    with torch.no_grad():   # 表示之后的运算不需要进行图构建
        y_pred = model.forward(X_test)
        y_pred_label = torch.where(y_pred >= 0.5, torch.tensor([1.0]), torch.tensor([0.0]))
        acc = torch.eq(y_pred_label, Y_test).sum().item() / Y_test.size(0)
        print("test acc: ", acc)


if __name__ == '__main__':
    for epoch in range(5):
        train(epoch)

    print('-------test-----------')
    test()