"""P2阶段: 预设KPI因果链（基于CMA四层因果链模型） 学习成长→内部流程→客户→财务 """ import sys, os sys.path.insert(0, os.path.dirname(os.path.dirname(__file__))) from app.database import get_engine from sqlalchemy import text # 基于CMA四层因果链模型的预设因果关系 PRESET_CAUSALITIES = [ # ===== 学习成长 → 内部流程 ===== # 培训完成率 → 过程合格率 {"source": "L_TRAINING", "target": "P_PASS_RATE", "strength": 0.6, "lag": 2, "direction": "positive", "formula": "培训提升技能→过程质量提升"}, # 人均培训时长 → 缺陷率（负相关） {"source": "L_TRAINING_HOURS", "target": "P_BUG_RATE", "strength": 0.5, "lag": 2, "direction": "negative", "formula": "更多培训→更少缺陷"}, # 员工满意度 → 返工率（负相关） {"source": "L_EMPLOYEE_SAT", "target": "P_REWORK_RATE", "strength": 0.4, "lag": 1, "direction": "negative", "formula": "员工满意→更少返工"}, # 关键技术掌握率 → 交付及时率 {"source": "L_TECH_COVERAGE", "target": "P_DELIVERY", "strength": 0.7, "lag": 1, "direction": "positive", "formula": "技术掌握→交付提升"}, # 关键岗位胜任度 → 过程合格率 {"source": "L_COMPETENCY", "target": "P_PASS_RATE", "strength": 0.6, "lag": 1, "direction": "positive", "formula": "胜任度→质量提升"}, # 创新提案数量 → 新产品收入占比 {"source": "L_INNOVATION_COUNT", "target": "P_NEW_PROD_RATIO", "strength": 0.5, "lag": 3, "direction": "positive", "formula": "创新提案→新产品上市"}, # 数据自动化率 → 交付及时率 {"source": "L_DATA_AUTO_RATE", "target": "P_DELIVERY", "strength": 0.3, "lag": 1, "direction": "positive", "formula": "自动化→效率提升"}, # 系统覆盖率 → 供应链响应周期 {"source": "L_SYS_COVERAGE", "target": "P_SUPPLY_CYCLE", "strength": 0.4, "lag": 3, "direction": "negative", "formula": "系统覆盖→周期缩短"}, # 战略认知度 → 新产品收入占比 {"source": "L_STRATEGY_AWARE", "target": "P_NEW_PROD_RATIO", "strength": 0.3, "lag": 2, "direction": "positive", "formula": "战略理解→创新聚焦"}, # ===== 内部流程 → 客户 ===== # 过程合格率 → 客户满意度 {"source": "P_PASS_RATE", "target": "C_SATISFACTION", "strength": 0.7, "lag": 1, "direction": "positive", "formula": "质量提升→客户满意"}, # 交付及时率 → 客户满意度 {"source": "P_DELIVERY", "target": "C_SATISFACTION", "strength": 0.6, "lag": 0, "direction": "positive", "formula": "及时交付→客户满意"}, # 缺陷率 → 客户满意度（负相关） {"source": "P_BUG_RATE", "target": "C_SATISFACTION", "strength": 0.5, "lag": 1, "direction": "negative", "formula": "缺陷多→不满意"}, # 返工率 → 客户满意度（负相关） {"source": "P_REWORK_RATE", "target": "C_SATISFACTION", "strength": 0.3, "lag": 1, "direction": "negative", "formula": "返工多→不满意"}, # 供应链响应周期 → 交付及时率（负相关） {"source": "P_SUPPLY_CYCLE", "target": "P_DELIVERY", "strength": 0.5, "lag": 1, "direction": "negative", "formula": "周期长→交付慢"}, # 产能利用率 → 交付及时率 {"source": "P_CAPACITY_UTIL", "target": "P_DELIVERY", "strength": 0.4, "lag": 0, "direction": "positive", "formula": "产能足→交付快"}, # 新产品收入占比 → 市场份额 {"source": "P_NEW_PROD_RATIO", "target": "C_MARKET_SHARE", "strength": 0.5, "lag": 3, "direction": "positive", "formula": "创新产品→市场占有率提升"}, # ===== 客户 → 财务 ===== # 客户满意度 → 客户保留率 {"source": "C_SATISFACTION", "target": "C_RETENTION_RATE", "strength": 0.8, "lag": 1, "direction": "positive", "formula": "满意→留存"}, # 客户满意度 → 营业收入 {"source": "C_SATISFACTION", "target": "F_REVENUE", "strength": 0.5, "lag": 2, "direction": "positive", "formula": "满意客户→复购增加"}, # 客户保留率 → 营业收入 {"source": "C_RETENTION_RATE", "target": "F_REVENUE", "strength": 0.6, "lag": 1, "direction": "positive", "formula": "老客户留存→稳定收入"}, # 客户保留率 → 客户生命周期价值 {"source": "C_RETENTION_RATE", "target": "C_LTV", "strength": 0.7, "lag": 2, "direction": "positive", "formula": "高留存→生命周期延长"}, # 新客户数 → 营业收入 {"source": "C_NEW_CLIENTS", "target": "F_REVENUE", "strength": 0.4, "lag": 1, "direction": "positive", "formula": "新客户→收入增长"}, # 市场份额 → 营业收入 {"source": "C_MARKET_SHARE", "target": "F_REVENUE", "strength": 0.5, "lag": 1, "direction": "positive", "formula": "市场扩大→收入增加"}, # 获客成本 → 净利润（负相关） {"source": "C_ACQUISITION_COST", "target": "F_NET_PROFIT", "strength": 0.3, "lag": 1, "direction": "negative", "formula": "获客成本高→利润减少"}, # 客户集中度 → 净利润（负相关） {"source": "C_CUST_CONCENTRATION", "target": "F_NET_PROFIT", "strength": 0.3, "lag": 1, "direction": "negative", "formula": "集中度高→风险增大"}, # ===== 财务内部因果 ===== # 营业收入 → 净利润 {"source": "F_REVENUE", "target": "F_NET_PROFIT", "strength": 0.7, "lag": 0, "direction": "positive", "formula": "收入增长→利润增加"}, # 毛利率 → 净利润 {"source": "F_GROSS_MARGIN", "target": "F_NET_PROFIT", "strength": 0.6, "lag": 0, "direction": "positive", "formula": "毛利提升→利润增加"}, # 费用率 → 净利润（负相关） {"source": "F_COST_RATIO", "target": "F_NET_PROFIT", "strength": 0.5, "lag": 0, "direction": "negative", "formula": "费用高→利润减少"}, # 收入增长率 → 营业收入 {"source": "F_REVENUE_GROWTH", "target": "F_REVENUE", "strength": 0.8, "lag": 1, "direction": "positive", "formula": "增长加速→收入提升"}, # 净利润 → 经济增加值 {"source": "F_NET_PROFIT", "target": "F_EVA", "strength": 0.9, "lag": 0, "direction": "positive", "formula": "净利润→经济增加值"}, # 经营性现金流 → 净利润（滞后反馈） {"source": "F_OP_CFLOW", "target": "F_NET_PROFIT", "strength": 0.4, "lag": 1, "direction": "positive", "formula": "现金充裕→运营改善"}, # 产品合格率 → 返工率（负相关） {"source": "F_QUALITY_RATE", "target": "F_REWORK_RATE", "strength": 0.6, "lag": 1, "direction": "negative", "formula": "合格率高→返工少"}, # 应收账款周转天数 → 经营性现金流（负相关） {"source": "F_AR_DAYS", "target": "F_OP_CFLOW", "strength": 0.5, "lag": 1, "direction": "negative", "formula": "回款慢→现金流紧张"}, # 流动比率 → 资产负债率 {"source": "F_CURRENT_RATIO", "target": "F_DEBT_RATIO", "strength": 0.3, "lag": 1, "direction": "negative", "formula": "流动性强→负债率低"}, # ===== 客户生命周期价值 ===== # 净推荐值(NPS) → 客户保留率 {"source": "C_NPS", "target": "C_RETENTION_RATE", "strength": 0.6, "lag": 1, "direction": "positive", "formula": "NPS高→留存好"}, # 客户生命周期价值 → 净利润 {"source": "C_LTV", "target": "F_NET_PROFIT", "strength": 0.5, "lag": 1, "direction": "positive", "formula": "客户终身价值→长期利润"}, ] def seed_causalities(): engine = get_engine() with engine.connect() as conn: # 先构建KPI代码→ID映射 kpis = conn.execute( text("SELECT id, kpi_code FROM kpi_definitions WHERE status='active'") ).fetchall() code_to_id = {r.kpi_code: r.id for r in kpis} inserted = 0 skipped = 0 for item in PRESET_CAUSALITIES: src_id = code_to_id.get(item["source"]) tgt_id = code_to_id.get(item["target"]) if not src_id: print(f" ⚠️ 源KPI不存在: {item['source']}") skipped += 1 continue if not tgt_id: print(f" ⚠️ 目标KPI不存在: {item['target']}") skipped += 1 continue # 检查是否已存在 existing = conn.execute( text("SELECT id FROM kpi_causality WHERE source_kpi_id=:src AND target_kpi_id=:tgt"), {"src": src_id, "tgt": tgt_id}, ).fetchone() if existing: skipped += 1 continue conn.execute( text("""INSERT INTO kpi_causality (source_kpi_id, target_kpi_id, strength, lag_months, formula, direction) VALUES (:src, :tgt, :strength, :lag, :formula, :dir)"""), { "src": src_id, "tgt": tgt_id, "strength": item["strength"], "lag": item["lag"], "formula": item["formula"], "dir": item["direction"], }, ) inserted += 1 conn.commit() print(f"因果链: 新增{inserted}, 跳过{skipped}") return inserted def seed_bi_templates(): engine = get_engine() PRESET_TEMPLATES = [ { "name": "四层指标总览", "report_type": "overview", "is_system": 1, "config": '{"description":"展示财务/客户/流程/学习四层维度KPI概览","layout":"grid","dimensions":["finance","customer","process","learning"],"metrics":["count","avg_value","alert_count"],"chart_type":"gauge_card"}', }, { "name": "同比趋势分析", "report_type": "trend", "is_system": 1, "config": '{"description":"各KPI近12个月趋势对比","period":"monthly","window_months":12,"chart_type":"line","show_compare":true}', }, { "name": "实际vs预算对比", "report_type": "comparison", "is_system": 1, "config": '{"description":"KPI实际值vs目标值偏差分析","chart_type":"bar","show_deviation":true,"group_by":"dimension"}', }, { "name": "TOP N异常KPI", "report_type": "topn", "is_system": 1, "config": '{"description":"排名前N的异常KPI红黄灯","top_n":10,"sort_by":"deviation","chart_type":"horizontal_bar","show_threshold":true}', }, { "name": "因果链推演", "report_type": "causality", "is_system": 1, "config": '{"description":"基于KPI因果链推演分析","chart_type":"force_graph","max_depth":3,"min_strength":0.3}', }, ] with engine.connect() as conn: inserted = 0 for tpl in PRESET_TEMPLATES: existing = conn.execute( text("SELECT id FROM bi_report_templates WHERE name=:name AND is_system=1"), {"name": tpl["name"]}, ).fetchone() if existing: continue conn.execute( text("INSERT INTO bi_report_templates (name, report_type, config, is_system) VALUES (:name, :type, :cfg, 1)"), {"name": tpl["name"], "type": tpl["report_type"], "cfg": tpl["config"]}, ) inserted += 1 conn.commit() print(f"报表模板: 新增{inserted}") return inserted if __name__ == "__main__": print("=== 种子数据初始化 ===\n") seed_causalities() seed_bi_templates() print("\n完成!")