划分训练集和验证集

划分训练集和验证集网上代码有很多，因为我在一个文件夹里面同时有image和label，如下图

 于是我参考了一篇知乎的文章，Pytorch将数据集划分为训练集、验证集和测试集 - 知乎在上面小做修改，就可以让数据集和验证集的图片+标签一起划分进入不同的文件夹。

import os
import random
import shutil
from shutil import copy2

def data_set_split(src_data_folder,src_data_label_folder,target_data_folder, train_scale=0.8, val_scale=0.2):
    print("开始数据集划分")
    # 在目标目录下创建文件夹
    split_names = ['train', 'val']
    for split_name in split_names:
        split_path = os.path.join(target_data_folder, split_name)
        if os.path.isdir(split_path):
            pass
        else:
            os.mkdir(split_path)
            
    current_class_data_path = os.path.join(src_data_folder)
    current_all_data = os.listdir(current_class_data_path)
    current_data_length = len(current_all_data)
    current_data_index_list = list(range(current_data_length))
    random.shuffle(current_data_index_list)
    
    current_class_data_label_path = os.path.join(src_data_label_folder)
    current_all_data_label = os.listdir(current_class_data_label_path)
    
    
    train_folder = os.path.join(os.path.join(target_data_folder, 'train'))   
    val_folder = os.path.join(os.path.join(target_data_folder, 'val'))
    train_stop_flag = current_data_length * train_scale
    val_stop_flag = current_data_length * val_scale                
    current_idx = 0
    train_num = 0
    val_num = 0
    
    for i in current_data_index_list:
        src_img_path = os.path.join(current_class_data_path, current_all_data[i])
        src_label_path = os.path.join(current_class_data_label_path, current_all_data_label[i])
        if current_idx < train_stop_flag:
            copy2(src_img_path, train_folder)
            copy2(src_label_path,train_folder)
                # print("{}复制到了{}".format(src_img_path, train_folder))
            train_num = train_num + 1
        else:
            copy2(src_img_path, val_folder)
            copy2(src_label_path,val_folder)
                # print("{}复制到了{}".format(src_img_path, val_folder))
            val_num = val_num + 1
        current_idx = current_idx + 1

if __name__ == '__main__':
    src_data_folder = "F:\\c_data\\images\\train"
    src_data_label_foler = "F:\\c_data\\labels\\train"
    target_data_folder = "F:\\training_dataset\\C"
    data_set_split(src_data_folder,src_data_label_foler, target_data_folder)

先将标签和图片放到两个文件夹里面src_data_folder、src_data_label_folder（这个直接windows里面分类复制粘贴就行）然后目标文件夹为target_data_folder里面分为训练集和验证集