Files
MoFin/scripts/data_validate.py
T
知微 9709c43ccb 总资产权威数据源统一修复
问题:总资产每次报告重新计算,数字不一致。
根因:cash字段错误(92664→73759),stale_push_wlin二次×0.866,
      报告各算各的。

修复:
1. portfolio.json cash 修正为Dad截图确认值73,758.85
2. price_monitor 每轮写入 total_mv + total_assets 到portfolio.json
   (从此所有报告只读这个字段,不自算)
3. stale_push_wlin 删除重复的 hmv *= 0.866(数据已CNY)
4. portfolio.json 加 currency: CNY 标记防混淆
5. 日志记录本次修复
2026-06-29 21:39:06 +08:00

90 lines
3.7 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
"""data_validate.py — 数据自检,在所有报告产出前执行
检查清单:
1. 总资产 = 市值 + 现金 (误差 < 1%
2. 持仓 vs 决策交叉检查
3. 币种一致性(港股必须currency=HKD
4. 数据时效:portfolio.json/decisions.json 今日已更新
返回值:通过→退出码0,输出"OK"。失败→退出码1,输出问题描述。
"""
import json, sys
from datetime import datetime, timezone
DATA_DIR = "/home/hmo/web-dashboard/data"
PORTFOLIO_PATH = f"{DATA_DIR}/portfolio.json"
DECISIONS_PATH = f"{DATA_DIR}/decisions.json"
STALE_REPORT = f"{DATA_DIR}/strategy_staleness_report.json"
issues = []
# ── 1. 总资产校验 ────────────────────────────────────────────
try:
pf = json.load(open(PORTFOLIO_PATH))
mv_calc = sum(h["shares"] * h["price"] for h in pf.get("holdings", []) if h.get("price"))
stored_ta = pf.get("total_assets", 0)
cash = pf.get("cash", 0)
expected_ta = round(mv_calc + cash, 2)
if stored_ta > 0 and abs(stored_ta - expected_ta) / stored_ta > 0.01:
issues.append(f"总资产不匹配: 存储{stored_ta} ≠ 计算{expected_ta} (市值{mv_calc}+现金{cash})")
except Exception as e:
issues.append(f"portfolio.json读取失败: {e}")
# ── 2. 持仓 vs 决策交叉检查 ──────────────────────────────────
try:
dec = json.load(open(DECISIONS_PATH))
dec_codes = {}
for d in dec.get("decisions", []):
dec_codes[d["code"]] = d
for h in pf.get("holdings", []):
code = h["code"]
if code not in dec_codes:
issues.append(f"持仓{code}({h.get('name','?')}) 在decisions.json中无对应决策")
for code, d in dec_codes.items():
if d.get("status") == "active" and d.get("type") == "持仓策略":
if not any(h["code"] == code for h in pf.get("holdings", [])):
issues.append(f"决策{code}({d.get('name','?')})标记持仓但portfolio.json中无此股")
except Exception as e:
issues.append(f"决策检查失败: {e}")
# ── 3. 币种一致性 ────────────────────────────────────────────
try:
for d in dec.get("decisions", []):
code = str(d.get("code", ""))
# 港股必须标记currency
if len(code) == 5 and code[0] in ("0", "1"):
cur = d.get("currency", "")
if cur not in ("HKD", "CNY"):
issues.append(f"港股{code}({d.get('name','?')}) 缺currency标记,不可靠")
# 如果标记了HKDstop_loss也应该是合理的HKD价(>10
sl = d.get("stop_loss", 0)
if cur == "HKD" and sl > 0 and sl < 1:
issues.append(f"港股{code} currency=HKD但stop_loss={sl} 异常低")
except Exception as e:
issues.append(f"币种检查失败: {e}")
# ── 4. 数据时效 ──────────────────────────────────────────────
today = datetime.now().strftime("%Y-%m-%d")
try:
if pf.get("updated_at", "").startswith(today):
pass # OK
else:
issues.append(f"portfolio.json updated_at={pf.get('updated_at','?')} 不是今日")
except:
pass
# ── 输出 ──────────────────────────────────────────────────────
if issues:
print("DATA_VALIDATE_FAIL")
for i in issues:
print(f" ⚠️ {i}")
sys.exit(1)
else:
print("DATA_VALIDATE_OK")
sys.exit(0)