fix: wire all factor runners into master.py and remove dead code

- master.py cmd_factors() now calls all 6 factor runner modules
- master.py INGEST_CHAIN includes ingest_events.py
- Remove dead compute_mom_season (O(N^2) Python loop, never registered)
- Fix unused variable in compute_mom12m_off_season

Author: yanpgwang

factors/predictors/price_extended_factors.py

@@ -126,21 +126,16 @@ def compute_momrev(df: pl.DataFrame) -> pl.DataFrame:
 def compute_mom12m_off_season(df: pl.DataFrame) -> pl.DataFrame:
     """12-month momentum excluding the same calendar month.
 
-    Standard Mom12m minus the return in the same month last year.
-    Approximated as: mom12m_cumret / (1 + ret_same_month_lag)
-    where ret_same_month_lag ≈ the return 252 trading days ago.
+    Approximated as: mom12m / (1 + ret_same_month_lag) - 1
     """
     panel = _ordered_panel(df, ["ts_code", "trade_date", "ret"])
     mom12 = ((1 + pl.col("ret")).log().rolling_sum(window_size=252).exp() - 1).shift(21)
-    # Approximate same-month return by single day 252 periods ago
     ret_season = pl.col("ret").shift(252)
-    expr = pl.when(ret_season.abs() < 0.99).then(
-        (1 + mom12) / (1 + ret_season) - 1
-    )
-    return _rolling_factor(panel.with_columns(
+    panel = panel.with_columns(
         mom12.over("ts_code").alias("_m12"),
         ret_season.over("ts_code").alias("_rs"),
-    ), "Mom12mOffSeason",
+    )
+    return _rolling_factor(panel, "Mom12mOffSeason",
         pl.when(pl.col("_rs").abs() < 0.99)
         .then((1 + pl.col("_m12")) / (1 + pl.col("_rs")) - 1)
     )

factors/predictors/price_volume_factors.py

@@ -152,55 +152,6 @@ def compute_mrreversal(df: pl.DataFrame) -> pl.DataFrame:
     return _rolling_factor(panel, "MRreversal", expr)
 
 
-# ---------------------------------------------------------------------------
-# Seasonal momentum (Heston-Sadka 2008 JFE): same-month avg over years 2-5
-# ---------------------------------------------------------------------------
-
-
-def compute_mom_season(df: pl.DataFrame) -> pl.DataFrame:
-    """Average return in the same calendar month over the preceding 2-5 years.
-
-    This requires monthly data. We compute it from the daily panel by first
-    collapsing to monthly returns, then averaging same-month returns from
-    years t-2 through t-5, then joining back to month-end dates.
-    """
-    panel = _ordered_panel(df, ["ts_code", "trade_date", "ret"])
-    panel = panel.with_columns(
-        (pl.col("trade_date").dt.year() * 100 + pl.col("trade_date").dt.month()).alias("yyyymm"),
-        pl.col("trade_date").dt.month().alias("_month"),
-        pl.col("trade_date").dt.year().alias("_year"),
-    )
-    monthly = (
-        panel.group_by(["ts_code", "yyyymm", "_month", "_year"])
-        .agg(
-            ((1 + pl.col("ret")).product() - 1).alias("month_ret"),
-            pl.col("trade_date").max().alias("trade_date"),
-        )
-        .sort(["ts_code", "yyyymm"])
-    )
-    rows = []
-    for (ts_code, month, year), sub in monthly.group_by(["ts_code", "_month", "_year"]):
-        same_month = monthly.filter(
-            (pl.col("ts_code") == ts_code)
-            & (pl.col("_month") == month)
-            & (pl.col("_year") >= year - 5)
-            & (pl.col("_year") <= year - 2)
-        )
-        if same_month.height == 0:
-            continue
-        avg = same_month["month_ret"].mean()
-        rows.append({"ts_code": ts_code, "yyyymm": year * 100 + month, "MomSeason": avg})
-
-    if not rows:
-        return pl.DataFrame(schema={"ts_code": pl.String, "trade_date": pl.Date, "MomSeason": pl.Float64})
-
-    result = pl.DataFrame(rows)
-    month_end = monthly.select(["ts_code", "yyyymm", "trade_date"])
-    return result.join(month_end, on=["ts_code", "yyyymm"], how="left").select(
-        ["ts_code", "trade_date", "MomSeason"]
-    )
-
-
 # ---------------------------------------------------------------------------
 # Log price (Blume-Husic 1973 JF)
 # ---------------------------------------------------------------------------

master.py

@@ -16,6 +16,7 @@ def _call(mod: str, fn: str = "run") -> None:
     "data.ingest.tushare.ingest_industry",
     "data.ingest.tushare.ingest_index",
     "data.ingest.tushare.ingest_status",
+    "data.ingest.tushare.ingest_events",
 ]
 
 NORMALIZE_CHAIN = [
@@ -53,6 +54,11 @@ def cmd_intermediate() -> None:
 
 def cmd_factors() -> None:
     _call("factors.predictors.run_seed_factors")
+    _call("factors.predictors.price_volume_factors")
+    _call("factors.predictors.accounting_factors")
+    _call("factors.predictors.extended_accounting_factors")
+    _call("factors.predictors.additional_accounting_factors")
+    _call("factors.predictors.price_extended_factors")
 
 
 def cmd_all() -> None: