在Python中，有多种方法可以导入训练集和测试集，以下是不同场景下的常见导入方法： ### 从CSV文件导入 如果数据集是以CSV文件形式存储的，可以使用`pandas`库来导入。结合引用[3]中的内容，示例代码如下： ```python import pandas as pd # 导入训练集 train_data = pd.read_csv("D:\\tmp\\ml_train.csv") # 导入测试集 test_data = pd.read_csv("D:\\tmp\\ml_test.csv") print("训练集数据信息：") train_data.info() print("测试集数据信息：") test_data.info() ``` ### 从图像和标签文件导入（适用于YOLO数据集） 对于YOLO模型的数据集，通常图像和标签文件是分开存储的。结合引用[1]，可以按以下步骤导入： ```python import os import cv2 # 定义训练集和测试集的图像与标签文件夹路径 train_image_folder = 'path/to/train/images' train_label_folder = 'path/to/train/labels' test_image_folder = 'path/to/test/images' test_label_folder = 'path/to/test/labels' # 导入训练集图像和标签 train_images = [] train_labels = [] for image_name in os.listdir(train_image_folder): image_path = os.path.join(train_image_folder, image_name) label_name = image_name.replace('.jpg', '.txt') # 假设标签文件扩展名是.txt label_path = os.path.join(train_label_folder, label_name) # 读取图像 image = cv2.imread(image_path) train_images.append(image) # 读取标签 with open(label_path, 'r') as f: label = f.read() train_labels.append(label) # 导入测试集图像和标签 test_images = [] test_labels = [] for image_name in os.listdir(test_image_folder): image_path = os.path.join(test_image_folder, image_name) label_name = image_name.replace('.jpg', '.txt') # 假设标签文件扩展名是.txt label_path = os.path.join(test_label_folder, label_name) # 读取图像 image = cv2.imread(image_path) test_images.append(image) # 读取标签 with open(label_path, 'r') as f: label = f.read() test_labels.append(label) ```