345 lines
17 KiB
Python
345 lines
17 KiB
Python
"""选股计划入池：把每日计划写进 Mongo 的股票池分组，供决策系统每晚推理覆盖。
|
||
|
||
背景（2026-08-03 与用户定稿）：决策系统 (bionic_trader) 每晚 22:30 的认知扫描，扫描
|
||
范围就是 Mongo `stock_groups` 集合里**所有分组的股票代码并集**。把计划写成一个独立
|
||
分组（默认 group_code=AKG_PLAN），候选票当晚就会被夜间推理覆盖，产出支撑位、压力位、
|
||
定性结论——持仓系统 (PMS) 的参考位、择时执行区间、研判上下文全部由此而来。
|
||
这就是「提前计算为主」：不新造任何推理步骤，把票放进池子，已有的夜间推理自然完成预计算。
|
||
|
||
入池、留池、出池的规则（用户拍板，decide() 的注释里逐条对应）：
|
||
入池 = 当日计划（强传导主榜前 N，与 PMS 候选同口径） ∪ 当前持仓
|
||
留池 = 旧成员既不在计划也无持仓、但形态未恶化的，留下继续接受每晚分析
|
||
——榜单是按条数截断过的，掉榜不等于变坏
|
||
出池 = 无持仓、不在计划、且形态已恶化（决策系统最新定性 SELL / AVOID / DROPPED）
|
||
→ 移入回收站集合 stock_recycle_bin（字段沿用现有格式，另加 reason 说明原因）
|
||
上限 = 池子超过 POOL_MAX 时，从「留池观察」里清最久没上过榜的（不进回收站；
|
||
出池后决策系统下次扫描会自动把其策略标 DROPPED 并出离场报告）
|
||
底线 = 持仓永不出池，即使形态恶化——那是 PMS 风控与体检的事，池子只保证它每晚有结论。
|
||
|
||
安全边界（都是「拿不到就不动」）：
|
||
* 计划数据缺失（因子表没跑出来）→ 整个入池动作中止，池子保持昨日原样。
|
||
* 持仓读不到 → 同样中止——读不到持仓就可能把持仓票错清出池，宁可不动。
|
||
* 决策系统结论读不到 → 本轮不判恶化、一只都不回收（缺数据不算恶化），其余照常。
|
||
|
||
用法：
|
||
python run.py push-pool --dry-run # 只打印入池/出池明细，不写库
|
||
python run.py push-pool # 真写（写完顺手触发决策系统的增量补扫）
|
||
python run.py push-pool --no-kick # 写库但不触发补扫（比如夜间已近 22:30 全量扫）
|
||
"""
|
||
import datetime as dt
|
||
import urllib.parse
|
||
import urllib.request
|
||
|
||
import common
|
||
import config
|
||
import db
|
||
import plan
|
||
|
||
|
||
# 决策系统结论里算「形态恶化」的定性（DROPPED 是它对掉出池子票的收尾标记）
|
||
BAD_SIGNALS = {"SELL", "AVOID", "DROPPED"}
|
||
|
||
|
||
# ============================================================================
|
||
# 纯逻辑：入池/留池/出池的决定（不碰任何库，test_pool_logic.py 直接测它）
|
||
# ============================================================================
|
||
def decide(old_members, member_meta, plan_codes, holdings, bad_codes,
|
||
max_size, today: str) -> dict:
|
||
"""按定稿规则算出新池子与各类进出明细。
|
||
|
||
old_members 上一版池子的代码集合
|
||
member_meta 上一版的成员记录 {code: {"added": 日期, "last_plan": 最近上榜日}}
|
||
plan_codes 当日计划代码（有序，榜单名次序）
|
||
holdings 当前持仓代码集合
|
||
bad_codes 形态已恶化的代码集合（判据 BAD_SIGNALS，缺数据时传空集=不回收）
|
||
"""
|
||
old = set(old_members or set())
|
||
plan_set = set(plan_codes or [])
|
||
hold = set(holdings or set())
|
||
meta = {k: dict(v) for k, v in (member_meta or {}).items()}
|
||
|
||
# 旧成员里既不在计划也无持仓的，按形态分流：恶化 → 回收站；未恶化 → 留池观察
|
||
idle = old - plan_set - hold
|
||
recycled = sorted(idle & set(bad_codes or set()))
|
||
observers = sorted(idle - set(recycled))
|
||
|
||
pool = list(dict.fromkeys(list(plan_codes or []) + sorted(hold) + observers))
|
||
|
||
# 上限：只清「留池观察」，按最久没上过榜的先清；计划与持仓永不清
|
||
cap_evicted = []
|
||
if max_size and len(pool) > max_size:
|
||
def _last_seen(c):
|
||
m = meta.get(c) or {}
|
||
return m.get("last_plan") or m.get("added") or ""
|
||
for c in sorted(observers, key=_last_seen):
|
||
if len(pool) <= max_size:
|
||
break
|
||
pool.remove(c)
|
||
cap_evicted.append(c)
|
||
observers = [c for c in observers if c not in cap_evicted]
|
||
|
||
# 成员记录：新进的记 added，今天在计划里的刷 last_plan，出池的删掉
|
||
for c in pool:
|
||
meta.setdefault(c, {"added": today})
|
||
if c in plan_set:
|
||
meta[c]["last_plan"] = today
|
||
for c in list(meta):
|
||
if c not in pool:
|
||
meta.pop(c)
|
||
|
||
return {
|
||
"pool": pool,
|
||
"new_entrants": sorted(plan_set - old), # 计划带来的新面孔
|
||
"retained_holdings": sorted(hold - plan_set), # 因持仓保留（不在当日计划里）
|
||
"observers": observers, # 留池观察
|
||
"recycled": recycled, # 移入回收站（形态恶化）
|
||
"cap_evicted": cap_evicted, # 池满出清（不进回收站）
|
||
"held_bad": sorted(hold & set(bad_codes or set())), # 持仓且形态恶化——只警示不出池
|
||
"meta": meta,
|
||
}
|
||
|
||
|
||
def build_remark(d: dict, plan_date: str, now_str: str, degraded: str = "") -> dict:
|
||
"""分组文档的 remark 字段，格式沿用现有池子的写法（summary / factor_details /
|
||
retained_positions / update_time_str），人读为主。"""
|
||
n_plan = len(d["pool"]) - len(d["retained_holdings"]) - len(d["observers"])
|
||
summary = (f"共入池 {len(d['pool'])} 只。当日计划 {n_plan} 只"
|
||
f"（其中新进 {len(d['new_entrants'])} 只），因持仓保留 "
|
||
f"{len(d['retained_holdings'])} 只，留池观察 {len(d['observers'])} 只；"
|
||
f"移入回收站 {len(d['recycled'])} 只（形态恶化），"
|
||
f"池满出清 {len(d['cap_evicted'])} 只。")
|
||
if d["held_bad"]:
|
||
summary += f" 警示：持仓中 {', '.join(d['held_bad'])} 形态已恶化（持仓不出池，请在 PMS 侧关注）。"
|
||
if degraded:
|
||
summary += f" 注意：{degraded}。"
|
||
plan_codes = [c for c in d["pool"]
|
||
if c not in set(d["retained_holdings"]) and c not in set(d["observers"])]
|
||
return {
|
||
"summary": summary,
|
||
"factor_details": [{
|
||
"factor_code": "akg_score", "trade_date": plan_date,
|
||
"selected_count": len(plan_codes), "selected_codes": plan_codes,
|
||
}],
|
||
"retained_positions": d["retained_holdings"],
|
||
"update_time_str": now_str,
|
||
}
|
||
|
||
|
||
def build_recycle_docs(d: dict, group_id: str, group_name: str, org_id: str,
|
||
now: dt.datetime) -> list:
|
||
"""回收站文档，字段照抄现有格式（group_id/group_name/org_id/removal_batch/
|
||
removed_at/stock_code），另加一个 reason 说明为什么移入——Mongo 加字段对
|
||
老读者无影响，但事后能分清「形态恶化」与其他原因。"""
|
||
batch = now.isoformat()
|
||
return [{"group_id": group_id, "group_name": group_name, "org_id": org_id,
|
||
"removal_batch": batch, "removed_at": now, "stock_code": c,
|
||
"reason": "形态恶化（决策系统最新定性 SELL/AVOID/DROPPED），且无持仓、不在当日计划"}
|
||
for c in d["recycled"]]
|
||
|
||
|
||
# ============================================================================
|
||
# 取数（每一路的失败语义见模块头「安全边界」）
|
||
# ============================================================================
|
||
def _read_holdings() -> set:
|
||
"""当前持仓。首选 PMS 账本 pms_position（新架构下它才是有人维护的持仓事实源：
|
||
下游 trading_position 表已没有写入方，2026-07-30 实测账户有仓时该表也是空的）；
|
||
账本表读不到再退回 trading_position 兜底。两条路都读不到就抛——持仓是出池判断
|
||
的底线输入，读不到宁可整轮不动池子。"""
|
||
try:
|
||
rows = db.read_mysql(
|
||
"pms", "SELECT ts_code, total_qty, status FROM pms_position")
|
||
out = set()
|
||
for r in rows.itertuples():
|
||
try:
|
||
qty = float(r.total_qty)
|
||
except (TypeError, ValueError):
|
||
qty = 0.0
|
||
if qty > 0 and str(r.status or "").upper() != "CLOSED":
|
||
out.add(common.to_prefix(str(r.ts_code).strip().upper()))
|
||
return out
|
||
except Exception as e: # noqa: BLE001 —— 账本表读不到才走下游表兜底
|
||
print(f" （PMS 账本 pms_position 读取失败，退回 trading_position 兜底: {e!r}）")
|
||
|
||
rows = db.read_mysql("pms", "SELECT * FROM trading_position")
|
||
if rows.empty:
|
||
return set()
|
||
cols = {c.lower(): c for c in rows.columns}
|
||
code_col = next((cols[c] for c in ("stock_code", "ts_code", "code") if c in cols), None)
|
||
qty_col = next((cols[c] for c in ("total_quantity", "current_qty", "total_qty",
|
||
"volume", "quantity") if c in cols), None)
|
||
if not code_col:
|
||
raise RuntimeError(f"trading_position 找不到代码列（现有列: {list(rows.columns)}）")
|
||
out = set()
|
||
for _, r in rows.iterrows():
|
||
code = str(r[code_col] or "").strip().upper()
|
||
if not code:
|
||
continue
|
||
try:
|
||
qty = float(r[qty_col]) if qty_col else 1.0
|
||
except (TypeError, ValueError):
|
||
qty = 1.0
|
||
if qty > 0:
|
||
out.add(common.to_prefix(code))
|
||
return out
|
||
|
||
|
||
def _read_bad_signals(codes: set):
|
||
"""这批票在决策系统结论表里的最新定性，恶化的挑出来。
|
||
读失败返回 (空集, 原因)——缺数据不算恶化，本轮不回收任何票。"""
|
||
if not codes:
|
||
return set(), ""
|
||
try:
|
||
marks = ",".join(["%s"] * len(codes))
|
||
rows = db.read_mysql(
|
||
"pms",
|
||
f"SELECT stock_code, signal_type, trade_date FROM strategy_daily_results "
|
||
f"WHERE stock_code IN ({marks})", tuple(codes))
|
||
except Exception as e: # noqa: BLE001
|
||
return set(), f"决策系统结论表读取失败（{e!r}），本轮不判恶化、不回收"
|
||
if rows.empty:
|
||
return set(), ""
|
||
rows = rows.sort_values("trade_date").drop_duplicates("stock_code", keep="last")
|
||
bad = {str(r.stock_code).strip().upper() for r in rows.itertuples()
|
||
if str(r.signal_type or "").strip().upper() in BAD_SIGNALS}
|
||
return bad, ""
|
||
|
||
|
||
def _mongo():
|
||
from pymongo import MongoClient
|
||
c = config.mongo()
|
||
uri = (f"mongodb://{urllib.parse.quote_plus(c.user)}:{urllib.parse.quote_plus(c.password)}"
|
||
f"@{c.host}:{c.port}/{c.db}?authSource=admin")
|
||
return MongoClient(uri, serverSelectionTimeoutMS=8000)
|
||
|
||
|
||
def _kick_bionic_scan() -> str:
|
||
"""写完池子后触发决策系统的增量补扫（只补当天还没分析过的票）。
|
||
失败只提示不报错——当晚 22:30 的全量扫描是兜底。"""
|
||
if not config.BIONIC_SCAN_URL or not config.BIONIC_SCAN_KEY:
|
||
return "未配置 BIONIC_SCAN_URL / BIONIC_SCAN_KEY，跳过补扫触发（当晚全量扫兜底）"
|
||
url = (f"{config.BIONIC_SCAN_URL}?key={urllib.parse.quote_plus(config.BIONIC_SCAN_KEY)}"
|
||
f"&mode=incremental")
|
||
try:
|
||
with urllib.request.urlopen(url, timeout=15) as resp:
|
||
body = resp.read().decode("utf-8", "replace")[:200]
|
||
return f"已触发决策系统增量补扫: {body}"
|
||
except Exception as e: # noqa: BLE001
|
||
return f"补扫触发失败（当晚 22:30 全量扫兜底）: {e!r}"
|
||
|
||
|
||
# ============================================================================
|
||
# 主流程
|
||
# ============================================================================
|
||
def push(date: str | None = None, top: int | None = None,
|
||
dry_run: bool = False, kick: bool = True) -> dict:
|
||
# 1. 当日计划（与 /plan 同一段装配代码）→ 先筛档位、再按绝对得分截断。
|
||
# 次序要紧（2026-08-17 修）：原来是"全档位取前 top、再筛强传导"——排进 top 的
|
||
# 弱/无传导票把 top 之外的强传导票挤掉了，与 PMS 候选的"先筛强传导再截断"分叉，
|
||
# 两边看到的候选不是同一批。所以这里向 collect 要一个宽池子（top 的十倍、至少
|
||
# 二百，全在打分池规模之内），先过档位白名单，最后取前 top。
|
||
# 主题限额默认 0（config.POOL_THEME_CAP，2026-08-17 拍板：分散不由选层做，
|
||
# 组合分散归 PMS 行业闸；配非零值即应急回退）。
|
||
top = top or config.POOL_TOP
|
||
data = plan.collect(date, top=max(top * 10, 200), obs_top=0,
|
||
theme_cap=config.POOL_THEME_CAP)
|
||
tiers = config.POOL_TIERS
|
||
plan_rows = [r for r in data["main"]
|
||
if not tiers or r.get("tier") in tiers][:top]
|
||
plan_codes = [r["code"] for r in plan_rows]
|
||
ds = data["date"]
|
||
|
||
# 2. 持仓（读不到直接抛，整轮不动池子）与旧池子
|
||
holdings = _read_holdings()
|
||
col_name = config.POOL_COLLECTION
|
||
client = None if dry_run and not config_mongo_ready() else _mongo()
|
||
old_doc, old_members, member_meta = None, set(), {}
|
||
if client is not None:
|
||
col = client[config.mongo().db][col_name]
|
||
old_doc = col.find_one({"group_code": config.POOL_GROUP_CODE,
|
||
"org_id": config.POOL_ORG_ID})
|
||
if old_doc:
|
||
old_members = {str(c).strip().upper() for c in old_doc.get("stock_codes") or []}
|
||
member_meta = old_doc.get("member_meta") or {}
|
||
|
||
# 3. 形态恶化名单（只查可能出池的那批；读失败=不回收）
|
||
idle = old_members - set(plan_codes) - holdings
|
||
bad, degraded = _read_bad_signals(idle | (holdings & old_members))
|
||
|
||
now = dt.datetime.now()
|
||
today = now.date().isoformat()
|
||
d = decide(old_members, member_meta, plan_codes, holdings, bad,
|
||
config.POOL_MAX, today)
|
||
|
||
# 4. 打印明细（干跑到此为止）
|
||
print(f"计划日 {ds}，档位白名单 {sorted(tiers) if tiers else '（不过滤）'}，"
|
||
f"计划入选 {len(plan_codes)} 只；持仓 {len(holdings)} 只；"
|
||
f"旧池 {len(old_members)} 只 → 新池 {len(d['pool'])} 只（上限 {config.POOL_MAX}）")
|
||
for label, items in (("计划新进", d["new_entrants"]),
|
||
("持仓保留", d["retained_holdings"]),
|
||
("留池观察", d["observers"]),
|
||
("移入回收站(形态恶化)", d["recycled"]),
|
||
("池满出清", d["cap_evicted"]),
|
||
("警示: 持仓且形态恶化(不出池)", d["held_bad"])):
|
||
if items:
|
||
print(f" {label} {len(items)} 只: {', '.join(items)}")
|
||
if degraded:
|
||
print(f" ⚠️ {degraded}")
|
||
if dry_run:
|
||
print("（--dry-run：只看不写）")
|
||
if client is not None:
|
||
client.close()
|
||
return d
|
||
|
||
# 5. 写分组文档（按 group_code+org_id 覆盖式更新）+ 回收站 + 触发补扫
|
||
remark = build_remark(d, ds, now.strftime("%Y-%m-%d %H:%M:%S"), degraded)
|
||
ctx = dict((old_doc or {}).get("strategy_context") or {})
|
||
ctx = {k: v for k, v in ctx.items() if k in set(d["pool"])}
|
||
for r in plan_rows:
|
||
ctx[r["code"]] = {"factor_code": "akg_score", "score": r.get("score"),
|
||
"tier": r.get("tier"), "upside": r.get("upside"),
|
||
"theme": (r.get("evidence") or {}).get("theme"),
|
||
"plan_date": ds}
|
||
for c in d["retained_holdings"]:
|
||
ctx.setdefault(c, {"factor_code": "akg_score", "note": "持仓保留"})
|
||
|
||
col = client[config.mongo().db][col_name]
|
||
col.update_one(
|
||
{"group_code": config.POOL_GROUP_CODE, "org_id": config.POOL_ORG_ID},
|
||
{"$set": {"group_name": config.POOL_GROUP_NAME, "pool_type": "core",
|
||
"is_public": False,
|
||
"description": "akg-factor-bridge 每日选股计划池：当日计划(强传导主榜) + "
|
||
"持仓保留 + 留池观察。决策系统每晚认知扫描按本组产出结论，"
|
||
"供 PMS 参考位/择时/研判使用。规则见 akg-factor-bridge/"
|
||
"docs/选股计划入池_对接说明.md",
|
||
"stock_codes": d["pool"], "member_meta": d["meta"],
|
||
"strategy_context": ctx, "remark": remark,
|
||
"updated_at": now},
|
||
"$setOnInsert": {"created_at": now}},
|
||
upsert=True)
|
||
doc = col.find_one({"group_code": config.POOL_GROUP_CODE,
|
||
"org_id": config.POOL_ORG_ID}, {"_id": 1})
|
||
print(f"✅ 分组已写入 {col_name}（group_code={config.POOL_GROUP_CODE}，"
|
||
f"_id={doc['_id']}，{len(d['pool'])} 只）")
|
||
|
||
if d["recycled"]:
|
||
bin_col = client[config.mongo().db][config.POOL_RECYCLE_COLLECTION]
|
||
docs = build_recycle_docs(d, str(doc["_id"]), config.POOL_GROUP_NAME,
|
||
config.POOL_ORG_ID, now)
|
||
bin_col.insert_many(docs)
|
||
print(f"✅ 回收站已记 {len(docs)} 只: {', '.join(d['recycled'])}")
|
||
client.close()
|
||
|
||
if kick:
|
||
print(_kick_bionic_scan())
|
||
return d
|
||
|
||
|
||
def config_mongo_ready() -> bool:
|
||
"""干跑时若 Mongo 还没配置（比如首次在开发机看效果），允许把旧池当空集。"""
|
||
try:
|
||
config.mongo()
|
||
return True
|
||
except Exception: # noqa: BLE001
|
||
print(" （Mongo 未配置，按空池试算）")
|
||
return False
|