#!/usr/bin/env python3 """ 收入完整性检测算法专项测试数据生成脚本 生成不同风险等级的测试场景数据 """ import os import sys import json import random from datetime import datetime, timedelta from typing import List, Dict, Any sys.path.append('/Users/liulujian/Documents/code/deeprisk-claude-1/backend') from loguru import logger logger.add("revenue_test_data.log", rotation="100 MB", level="INFO") class RevenueTestDataGenerator: """收入完整性检测专项测试数据生成器""" def __init__(self): self.output_dir = "/Users/liulujian/Documents/code/deeprisk-claude-1/backend/test_data/revenue_test" os.makedirs(self.output_dir, exist_ok=True) def generate_streamer_info(self) -> List[Dict]: """生成主播信息数据""" logger.info("生成主播信息数据...") streamers = [ { "streamer_id": "TEST_001", "streamer_name": "测试主播-严重漏报", "tax_no": "91110000123456789A", "platform": "抖音", "tier": "S", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_002", "streamer_name": "测试主播-高风险", "tax_no": "91110000987654321B", "platform": "快手", "tier": "A", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_003", "streamer_name": "测试主播-中风险", "tax_no": "91110000555666777C", "platform": "淘宝直播", "tier": "B", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_004", "streamer_name": "测试主播-低风险", "tax_no": "91110000333444555D", "platform": "小红书", "tier": "A", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_005", "streamer_name": "测试主播-正常", "tax_no": "91110000111222333E", "platform": "抖音", "tier": "S", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_006", "streamer_name": "测试主播-完全未申报", "tax_no": "91110000777888999F", "platform": "快手", "tier": "S", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_007", "streamer_name": "测试主播-多平台收入", "tax_no": "91110000222333444G", "platform": "抖音", "tier": "S", "status": "active", "created_at": "2024-01-01 00:00:00" }, { "streamer_id": "TEST_008", "streamer_name": "测试主播-分批申报", "tax_no": "91110000555666777H", "platform": "快手", "tier": "A", "status": "active", "created_at": "2024-01-01 00:00:00" } ] logger.info(f"成功生成 {len(streamers)} 个主播信息") return streamers def generate_revenue_sharing_contracts(self, streamers: List[Dict]) -> List[Dict]: """生成主播分成协议数据""" logger.info("生成分成协议数据...") contracts = [] base_date = datetime(2024, 1, 1) for streamer in streamers: # 为每个主播创建多个分成协议 for i in range(3): # 每个主播3个协议 contract = { "contract_id": f"CONTRACT_{streamer['streamer_id']}_{i+1:03d}", "streamer_id": streamer['streamer_id'], "platform": streamer['platform'], "start_date": (base_date + timedelta(days=i*30)).strftime("%Y-%m-%d"), "end_date": (base_date + timedelta(days=365)).strftime("%Y-%m-%d"), "share_ratio": round(random.uniform(0.3, 0.7), 2), # 30%-70%分成 "platform_share_ratio": round(random.uniform(0.3, 0.7), 2), # 平台分成 "status": "active", "created_at": (base_date + timedelta(days=i*30)).strftime("%Y-%m-%d %H:%M:%S") } contracts.append(contract) logger.info(f"成功生成 {len(contracts)} 个分成协议") return contracts def generate_recharge_data(self, streamers: List[Dict]) -> List[Dict]: """生成充值数据 - 不同风险场景""" logger.info("生成充值数据（多风险场景）...") recharges = [] base_date = datetime(2024, 1, 1) # 场景1: 严重漏报 (TEST_001) - 充值50万，申报10万 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_001') for day in range(1, 31): # 30天 amount = random.randint(10000, 20000) # 每天1-2万 recharges.append({ "recharge_id": f"RECH_001_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景2: 高风险 (TEST_002) - 充值30万，申报18万 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_002') for day in range(1, 31): amount = random.randint(8000, 12000) recharges.append({ "recharge_id": f"RECH_002_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景3: 中风险 (TEST_003) - 充值10万，申报8.5万 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_003') for day in range(1, 31): amount = random.randint(2500, 4000) recharges.append({ "recharge_id": f"RECH_003_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景4: 低风险 (TEST_004) - 充值5万，申报4.7万 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_004') for day in range(1, 31): amount = random.randint(1200, 2000) recharges.append({ "recharge_id": f"RECH_004_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景5: 正常 (TEST_005) - 充值10万，申报9.8万（正常误差） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_005') for day in range(1, 31): amount = random.randint(2500, 4000) recharges.append({ "recharge_id": f"RECH_005_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景6: 完全未申报 (TEST_006) - 充值80万，申报0 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_006') for day in range(1, 31): amount = random.randint(20000, 30000) recharges.append({ "recharge_id": f"RECH_006_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景7: 多平台收入 (TEST_007) - 充值15万，申报5万 streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_007') for day in range(1, 31): amount = random.randint(4000, 6000) recharges.append({ "recharge_id": f"RECH_007_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) # 场景8: 分批申报 (TEST_008) - 充值12万，申报6万（分批） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_008') for day in range(1, 31): amount = random.randint(3000, 5000) recharges.append({ "recharge_id": f"RECH_008_{day:03d}", "streamer_id": streamer['streamer_id'], "recharge_date": (base_date + timedelta(days=day-1)).strftime("%Y-%m-%d"), "recharge_amount": float(amount), "payment_method": random.choice(["支付宝", "微信支付", "银行卡"]), "payment_status": "completed", "platform": streamer['platform'] }) logger.info(f"成功生成 {len(recharges)} 条充值记录") return recharges def generate_tax_declarations(self, streamers: List[Dict]) -> List[Dict]: """生成税务申报数据 - 与充值数据对应的申报""" logger.info("生成税务申报数据...") declarations = [] base_date = datetime(2024, 1, 31) # 月末申报 # 场景1: 严重漏报 - 申报10万（实际充值约50万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_001') declarations.append({ "declaration_id": f"TAX_001_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 100000.0, # 只申报10万 "tax_rate": 0.13, "tax_amount": 13000.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景2: 高风险 - 申报18万（实际充值约30万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_002') declarations.append({ "declaration_id": f"TAX_002_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 180000.0, "tax_rate": 0.13, "tax_amount": 23400.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景3: 中风险 - 申报8.5万（实际充值约10万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_003') declarations.append({ "declaration_id": f"TAX_003_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 85000.0, "tax_rate": 0.13, "tax_amount": 11050.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景4: 低风险 - 申报4.7万（实际充值约5万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_004') declarations.append({ "declaration_id": f"TAX_004_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 47000.0, "tax_rate": 0.13, "tax_amount": 6110.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景5: 正常 - 申报9.8万（实际充值约10万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_005') declarations.append({ "declaration_id": f"TAX_005_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 98000.0, "tax_rate": 0.13, "tax_amount": 12740.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景6: 完全未申报 - 申报0（实际充值约80万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_006') declarations.append({ "declaration_id": f"TAX_006_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 0.0, # 完全未申报 "tax_rate": 0.13, "tax_amount": 0.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景7: 多平台收入 - 申报5万（实际充值约15万） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_007') declarations.append({ "declaration_id": f"TAX_007_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 50000.0, "tax_rate": 0.13, "tax_amount": 6500.0, "declaration_period": "2024-01", "status": "submitted" }) # 场景8: 分批申报 - 申报6万（实际充值约12万，分批只申报一半） streamer = next(s for s in streamers if s['streamer_id'] == 'TEST_008') declarations.append({ "declaration_id": f"TAX_008_2024-01", "tax_no": streamer['tax_no'], "streamer_id": streamer['streamer_id'], "declaration_date": base_date.strftime("%Y-%m-%d"), "declared_amount": 60000.0, "tax_rate": 0.13, "tax_amount": 7800.0, "declaration_period": "2024-01", "status": "submitted" }) logger.info(f"成功生成 {len(declarations)} 条税务申报记录") return declarations def generate_test_scenarios(self) -> Dict[str, Dict]: """生成测试场景说明""" scenarios = { "scenario_1": { "name": "严重漏报场景", "streamer_id": "TEST_001", "expected_risk_level": "CRITICAL", "expected_risk_score": "90-100", "description": "充值50万，仅申报10万，差异率80%，差异金额40万", "test_focus": "测试严重风险检测能力" }, "scenario_2": { "name": "高风险场景", "streamer_id": "TEST_002", "expected_risk_level": "HIGH", "expected_risk_score": "75-90", "description": "充值30万，申报18万，差异率40%，差异金额12万", "test_focus": "测试高风险检测能力" }, "scenario_3": { "name": "中风险场景", "streamer_id": "TEST_003", "expected_risk_level": "MEDIUM", "expected_risk_score": "50-75", "description": "充值10万，申报8.5万，差异率15%，差异金额1.5万", "test_focus": "测试中风险检测能力" }, "scenario_4": { "name": "低风险场景", "streamer_id": "TEST_004", "expected_risk_level": "LOW", "expected_risk_score": "25-50", "description": "充值5万，申报4.7万，差异率6%，差异金额3000元", "test_focus": "测试低风险检测能力" }, "scenario_5": { "name": "正常场景", "streamer_id": "TEST_005", "expected_risk_level": "LOW", "expected_risk_score": "0-25", "description": "充值10万，申报9.8万，差异率2%，差异金额2000元", "test_focus": "测试正常数据处理能力" }, "scenario_6": { "name": "完全未申报场景", "streamer_id": "TEST_006", "expected_risk_level": "CRITICAL", "expected_risk_score": "95-100", "description": "充值80万，申报0，差异率100%，差异金额80万", "test_focus": "测试极端情况检测能力" }, "scenario_7": { "name": "多平台收入场景", "streamer_id": "TEST_007", "expected_risk_level": "HIGH", "expected_risk_score": "80-90", "description": "充值15万，申报5万，差异率67%，差异金额10万", "test_focus": "测试多平台收入检测能力" }, "scenario_8": { "name": "分批申报场景", "streamer_id": "TEST_008", "expected_risk_level": "MEDIUM", "expected_risk_score": "60-75", "description": "充值12万，申报6万，差异率50%，差异金额6万", "test_focus": "测试分批申报检测能力" } } return scenarios def save_data(self, data: Any, filename: str): """保存数据到文件""" filepath = os.path.join(self.output_dir, filename) with open(filepath, 'w', encoding='utf-8') as f: json.dump(data, f, ensure_ascii=False, indent=2, default=str) logger.info(f"数据已保存到: {filepath}") def generate_all(self): """生成所有测试数据""" logger.info("开始生成收入完整性检测专项测试数据...") # 1. 生成基础数据 streamers = self.generate_streamer_info() contracts = self.generate_revenue_sharing_contracts(streamers) recharges = self.generate_recharge_data(streamers) declarations = self.generate_tax_declarations(streamers) scenarios = self.generate_test_scenarios() # 2. 保存数据 self.save_data(streamers, "streamers.json") self.save_data(contracts, "contracts.json") self.save_data(recharges, "recharges.json") self.save_data(declarations, "tax_declarations.json") self.save_data(scenarios, "test_scenarios.json") # 3. 生成汇总报告 summary = { "generation_time": datetime.now().isoformat(), "description": "收入完整性检测算法专项测试数据集", "data_count": { "streamers": len(streamers), "contracts": len(contracts), "recharges": len(recharges), "tax_declarations": len(declarations), "test_scenarios": len(scenarios) }, "total_records": len(streamers) + len(contracts) + len(recharges) + len(declarations), "risk_scenarios": list(scenarios.keys()) } self.save_data(summary, "summary.json") # 4. 打印统计信息 logger.info("=" * 60) logger.info("收入完整性检测专项测试数据生成完成！") logger.info("=" * 60) logger.info(f"总计生成 {summary['total_records']} 条记录") logger.info(f"包含 {len(scenarios)} 个测试场景") logger.info("") logger.info("测试场景：") for scenario_id, scenario in scenarios.items(): logger.info(f" {scenario_id}: {scenario['name']} - {scenario['expected_risk_level']}") # 5. 生成统计表 self.print_statistics(recharges, declarations, scenarios) return summary def print_statistics(self, recharges: List[Dict], declarations: List[Dict], scenarios: Dict): """打印详细统计信息""" print("\n" + "=" * 80) print("收入完整性检测算法测试数据统计表") print("=" * 80) for scenario_id, scenario in scenarios.items(): streamer_id = scenario['streamer_id'] # 计算该主播的充值总额 streamer_recharges = [r for r in recharges if r['streamer_id'] == streamer_id] recharge_total = sum(r['recharge_amount'] for r in streamer_recharges) # 获取对应的申报数据 declaration = next((d for d in declarations if d['streamer_id'] == streamer_id), None) declared_amount = declaration['declared_amount'] if declaration else 0 # 计算差异 difference = recharge_total - declared_amount difference_rate = (difference / recharge_total * 100) if recharge_total > 0 else 0 print(f"\n场景 {scenario_id}: {scenario['name']}") print(f" 预期风险等级: {scenario['expected_risk_level']}") print(f" 充值总额: ¥{recharge_total:,.2f}") print(f" 申报金额: ¥{declared_amount:,.2f}") print(f" 差异金额: ¥{difference:,.2f}") print(f" 差异率: {difference_rate:.2f}%") def main(): """主函数""" print("=" * 80) print("收入完整性检测算法 - 专项测试数据生成器") print("=" * 80) print() print("本工具将为收入完整性检测算法生成8个不同风险等级的测试场景：") print(" 1. 严重漏报 (CRITICAL)") print(" 2. 高风险 (HIGH)") print(" 3. 中风险 (MEDIUM)") print(" 4. 低风险 (LOW)") print(" 5. 正常 (LOW)") print(" 6. 完全未申报 (CRITICAL)") print(" 7. 多平台收入 (HIGH)") print(" 8. 分批申报 (MEDIUM)") print("=" * 80) print() generator = RevenueTestDataGenerator() try: generator.generate_all() print("\n✅ 测试数据生成成功！") print(f"📂 数据位置: /Users/liulujian/Documents/code/deeprisk-claude-1/backend/test_data/revenue_test/") print("\n📋 使用方法:") print(" 1. 查看测试场景: cat test_scenarios.json") print(" 2. 查看统计数据: cat summary.json") print(" 3. 运行算法测试: 使用 streamer_id 进行测试") print("\n🔬 API测试示例:") print(' curl -X POST http://localhost:8000/api/v1/detect \\') print(' -H "Content-Type: application/json" \\') print(' -d \'{"streamer_id": "TEST_001", "period": "2024-01"}\'') except Exception as e: logger.error(f"生成测试数据失败: {str(e)}", exc_info=True) print(f"\n❌ 生成失败: {str(e)}") sys.exit(1) if __name__ == "__main__": main()