PaddlePaddle 自定义回归预测数据集

记录PaddlePaddle继承Dataset类实现自定义回归预测数据集的相关代码（Pytorch类似）

import paddle
from paddle.io import Dataset
from sklearn.model_selection import train_test_split

class MyDataset(Dataset):
    def __init__(self, data,mode='train'):

        super().__init__()
        train_ratio = 0.8
        test_ratio = 0.2
        self.data = data
        self.features = list(self.data.columns)[2:-1]
        self.label = list(self.data.columns)[-1]
        self.data = self.data.dropna()
        self.train_data,self.test_data = train_test_split(self.data,
                                                          train_size=train_ratio,
                                                          test_size=test_ratio,
                                                          shuffle=False)
        if mode == 'train':
            self.data = self.train_data
        if mode == 'test':
            self.data = self.test_data

    def __getitem__(self, index):

        if index >= len(self):
            raise ValueError("Index out of range")
        record = self.data.iloc[index]   #根据索引值指定行
        features = [record[feature]for feature in self.features]
        label = record[self.label]
        features = paddle.to_tensor(features,dtype='float32')
        label = paddle.to_tensor(label,dtype='float32')

        return features,label

    def __len__(self):

        return len(self.data)

if __name__ == '__main__':
    # 测试定义的数据集
    data = pd.read_excel('./dataset/result.xlsx')
    train_dataset = MyDataset(data, mode='train')
    test_dataset = MyDataset(data, mode='test')
    train_loader = paddle.io.DataLoader(train_dataset,batch_size=16,shuffle=False)
    test_loader = paddle.io.DataLoader(test_dataset,batch_size=16,shuffle=False)

所测试数据集如下图所示：

此处需要指定特征及标签列