```python # 假设你要创建并保存这三个文件的模拟数据，以下是对应的代码： import pandas as pd import numpy as np from datetime import datetime, timedelta import random # 设置随机种子 np.random.seed(42) random.seed(42) # 参数 n_users = 1000 n_orders = 3000 n_browses = 10000 categories = ['电子产品', '服装', '家居用品', '图书', '美妆', '食品'] regions = ['华东', '华北', '华南', '华中', '西南', '西北', '东北'] genders = ['男', '女'] # ========== 生成 user_info.csv ========== user_info = pd.DataFrame({ 'user_id': range(1, n_users + 1), 'age': np.random.randint(18, 65, size=n_users), 'gender': np.random.choice(genders, size=n_users), 'region': np.random.choice(regions, size=n_users), 'register_date': [datetime(2023, 1, 1) + timedelta(days=random.randint(0, 365)) for _ in range(n_users)] }) user_info.to_csv('user_info.csv', index=False, encoding='utf-8') print("✅ user_info.csv 已生成：用户基本信息") # ========== 生成 order_data.csv ========== def random_date(): start = datetime(2023, 1, 1) return start + timedelta(days=random.randint(0, 365), hours=random.randint(0, 23)) order_data = [] for i in range(n_orders): user = user_info.sample(1).iloc[0] purchase_date = random_date() amount = round(np.random.lognormal(mean=3.5, sigma=1.0), 2) order_data.append({ 'order_id': i + 1, 'user_id': user['user_id'], 'purchase_date': purchase_date, 'category': random.choice(categories), 'amount': max(amount, 5), 'quantity': np.random.randint(1, 5) }) order_data = pd.DataFrame(order_data).sort_values('purchase_date').reset_index(drop=True) order_data.to_csv('order_data.csv', index=False, encoding='utf-8') print("✅ order_data.csv 已生成：订单购买记录") # ========== 生成 browse_data.csv ========== browse_data = [] for i in range(n_browses): user = user_info.sample(1).iloc[0] view_time = random_date() browse_data.append({ 'browse_id': i + 1, 'user_id': user['user_id'], 'view_time': view_time, 'page': f"product_{random.randint(1, 200)}", 'category': random.choice(categories), 'duration_seconds': round(np.random.exponential(scale=90), 2), 'pages_viewed': np.random.poisson(lam=4) }) browse_data = pd.DataFrame(browse_data).sort_values('view_time').reset_index(drop=True) browse_data.to_csv('browse_data.csv', index=False, encoding='utf-8') print("✅ browse_data.csv 已生成：用户浏览行为") # 输出前几行示例 print("\n📄 user_info 前5行：") print(user_info.head()) print("\n📄 order_data 前5行：") print(order_data.head()) print("\n📄 browse_data 前5行：") print(browse_data.head()) ```