finish experiment 47 (exp/47-q16-adding-hmm-regime-posterior-features)

code/MANIFEST.txt

@@ -1,19 +1,31 @@
 # TradeAC custom-qlib-code snapshot (auto-generated)
-# parent repo HEAD : 71211552910ee2ebaed853696e9ad8d89641c3de
+# parent repo HEAD : adc685d9cfe7556ef314cc9dc7c7bccb482aa2f8
 #   tac-qlib/tac_qlib/contrib
 #   tac-qlib/tac_qlib/data
 # per-file hashes (git hash-object):
   1b6298c4a5652f2e863cbdc385a1014a570fcd59  tac-qlib/tac_qlib/contrib/__init__.py
+  b419ee55ed455a1c45423d1c9025ca5cc0a98576  tac-qlib/tac_qlib/contrib/__pycache__/__init__.cpython-312.pyc
   c76a9f17f680e74eea766eff27f7624359749ed6  tac-qlib/tac_qlib/contrib/data/__init__.py
-  871ff1e163c29261f140c3f53d42a41e6504c779  tac-qlib/tac_qlib/contrib/data/handler.py
+  2f6c67620aa2f9e6aaaef3369361d9b3eac3d6ca  tac-qlib/tac_qlib/contrib/data/__pycache__/__init__.cpython-312.pyc
+  fdd5923a70a399e8680913593ff111641947898e  tac-qlib/tac_qlib/contrib/data/__pycache__/handler.cpython-312.pyc
+  0dd25ef161c6e0f15eafc84886e7e1381deb38c3  tac-qlib/tac_qlib/contrib/data/handler.py
   b151d139a0dcde87d74b21e7c4b729176ba5c39b  tac-qlib/tac_qlib/contrib/model/__init__.py
+  08dec87ccdf6bb5d2cf611ca3032a4280aaab8cf  tac-qlib/tac_qlib/contrib/model/__pycache__/__init__.cpython-312.pyc
+  6fb61946ea9a83dfb560de3717f5fbf482c4c00e  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_ensemble.cpython-312.pyc
+  3e80f2e08b661ddd2f58ffe5a6196063fa41ae51  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_gbdt.cpython-312.pyc
   d3f051f3a8650c42fedc7b367b966f7c74fb5789  tac-qlib/tac_qlib/contrib/model/rank_ensemble.py
-  ccfe7d554989aa7f3e5a2128ae663e51b2207149  tac-qlib/tac_qlib/contrib/model/rank_gbdt.py
+  d03e6611338918d4aac5eea4adf26f85a3763652  tac-qlib/tac_qlib/contrib/model/rank_gbdt.py
-  4afcf9058231111c412925f4c4b84e81d656db87  tac-qlib/tac_qlib/contrib/strategy/__init__.py
+  c4ef84ffda2a611262412fe1127689c667f3d0c1  tac-qlib/tac_qlib/contrib/strategy/__init__.py
+  6ad10c2ebe37c16417e67c7aeb731ad1fcb6da2f  tac-qlib/tac_qlib/contrib/strategy/__pycache__/__init__.cpython-312.pyc
+  8d684b3216b040071d9ee4fa920a0e0c7486d278  tac-qlib/tac_qlib/contrib/strategy/__pycache__/optimal_stop.cpython-312.pyc
+  896ef74ae47bcd1ed388e1e5d9c8d70c28097fe9  tac-qlib/tac_qlib/contrib/strategy/kelly_dropout.py
   79aaad9e39fcc740a773f4f63c512ce1086cfde0  tac-qlib/tac_qlib/contrib/strategy/optimal_stop.py
+  5b9acfb4340111b204249add7760bd53c6ae03f1  tac-qlib/tac_qlib/contrib/strategy/regime_gate.py
+  aa1ee880d52ceb5821d65973962099c2254f710a  tac-qlib/tac_qlib/contrib/strategy/top_bottom.py
+  fe60bacdfedd48617863be31f24b7c7daebfac5a  tac-qlib/tac_qlib/contrib/strategy/weekly_rebalance.py
   92e6e90eb0cd0a25142034560f27adb6b705b1a8  tac-qlib/tac_qlib/data/__init__.py
-  0eefe65b3210b764ebf52ecd27e405c4b5ecaa8f  tac-qlib/tac_qlib/data/__pycache__/__init__.cpython-312.pyc
+  7c4e6c345fad1978efe8860c0d977d0c02d6f8d9  tac-qlib/tac_qlib/data/__pycache__/__init__.cpython-312.pyc
-  4d5f85b06931f153942870c922f13713dd632ace  tac-qlib/tac_qlib/data/__pycache__/config.cpython-312.pyc
+  99e602392d51663cb06d5c425000b1ed1e5a916b  tac-qlib/tac_qlib/data/__pycache__/config.cpython-312.pyc
-  158481e2d49653b6d8c71d88dc4ba3ac1148c57a  tac-qlib/tac_qlib/data/__pycache__/providers.cpython-312.pyc
+  020dcdcf288e4832c8cf2386351f78d5ceb4fe13  tac-qlib/tac_qlib/data/__pycache__/providers.cpython-312.pyc
-  686d36f6d101c547491ca866aa143aa542e17518  tac-qlib/tac_qlib/data/config.py
+  53c9007a928841fd3c3b08450f9a6520ce1ac091  tac-qlib/tac_qlib/data/config.py
-  d9f839be30026f337754a3f015425a8efdbe8e2a  tac-qlib/tac_qlib/data/providers.py
+  8d0644f6f0d1efb94798ed444cc73e63b643459b  tac-qlib/tac_qlib/data/providers.py

code/tac-qlib/tac_qlib/contrib/data/handler.py

@@ -64,9 +64,13 @@ def check_transform_proc(proc_l, fit_start_time, fit_end_time):
 
 
 def get_common_feature_fields(lake_root=None, market="US", timeframe="1d") -> List[str]:
-    """Discover ta-lib columns present in *every* features parquet file of the lake.
+    """Discover feature columns present in *every* feature file of the lake.
 
-    Returns sorted field names (without the ``$`` prefix). Empty if no features are persisted.
+    Walks the `family=ta|sp` partition layout (plus any legacy flat files).
+    TA and SP columns are disjoint by construction, so the common set is
+    computed per family (columns shared by all symbol files of that family),
+    then the per-family results are unioned. Returns sorted field names
+    (without the ``$`` prefix). Empty if no features are persisted.
     """
     cfg = LakeConfig(lake_root, market)
     feat_dir = cfg.features_dir(timeframe)
@@ -74,8 +78,9 @@ def get_common_feature_fields(lake_root=None, market="US", timeframe="1d") -> Li
         return []
     import pyarrow.parquet as pq
 
+    def _family_common(fam_dir: Path) -> set:
         common = None
-    for p in sorted(feat_dir.glob("symbol=*.parquet")):
+        for p in sorted(fam_dir.glob("symbol=*.parquet")):
             try:
                 cols = set(pq.read_schema(p).names) - set(NON_FEATURE_COLUMNS)
             except Exception:  # pragma: no cover - skip unreadable files
@@ -83,7 +88,20 @@ def get_common_feature_fields(lake_root=None, market="US", timeframe="1d") -> Li
             common = cols if common is None else (common & cols)
             if not common:
                 break
-    return sorted(common) if common else []
+        return common or set()
+
+    common: set = set()
+    # family tier: features/market=*/timeframe=*/family=*/symbol=*.parquet
+    for fam in ("ta", "sp"):
+        fam_dir = feat_dir / f"family={fam}"
+        if fam_dir.is_dir():
+            common |= _family_common(fam_dir)
+    # legacy flat: features/market=*/timeframe=*/symbol=*.parquet
+    if (feat_dir / "family=ta").exists() or (feat_dir / "family=sp").exists():
+        pass  # family layout already covered
+    else:
+        common |= _family_common(feat_dir)
+    return sorted(common)
 
 
 class DropAllNaN(processor_module.Processor):

code/tac-qlib/tac_qlib/contrib/model/rank_gbdt.py

@@ -53,23 +53,61 @@ from qlib.workflow import R
 __all__ = ["RankICLGBModel", "rankic_feval"]
 
 
+def _group_averaged_rank(values: np.ndarray, gid: np.ndarray, offs: np.ndarray) -> np.ndarray:
+    """Averaged (tie-corrected) rank of ``values`` within each group, vectorized.
+
+    ``gid`` maps each row to its group id; ``offs`` holds the cumulative row
+    offsets so that group ``i`` occupies rows ``[offs[i], offs[i+1])``. Returns
+    the same result as ``pandas.Series.rank(method='average')`` applied per
+    group, but in one pass (``np.lexsort`` is the only non-linear step).
+    """
+    n = len(values)
+    order = np.lexsort((values, gid))
+    ord_rank = np.empty(n, dtype=np.float64)
+    ord_rank[order] = np.arange(n, dtype=np.float64) - offs[gid[order]] + 1.0
+    sg = gid[order]
+    sv = values[order]
+    newblock = np.empty(n, dtype=bool)
+    newblock[0] = True
+    newblock[1:] = (sg[1:] != sg[:-1]) | (sv[1:] != sv[:-1])
+    blockid = np.cumsum(newblock) - 1
+    block_mean = np.bincount(blockid, weights=ord_rank[order]) / np.bincount(blockid)
+    out = np.empty(n)
+    out[order] = block_mean[blockid]
+    return out
+
+
 def _per_day_spearman(preds: np.ndarray, labels: np.ndarray, group: np.ndarray) -> float:
     """Mean per-day Spearman rank correlation of preds vs labels.
 
     ``group`` holds the number of rows of each trading day (query group), in
     order. Days with <3 valid rows or a constant pred/label are skipped.
+
+    Vectorized: per-day Spearman == Pearson of the per-day rank transforms,
+    and the Pearson moments (``sum``, ``sum`` of products/squares) aggregate
+    over each day with ``np.bincount``. Runs ~10x faster than the per-day
+    ``pd.Series.rank()`` loop that preceded it — this feval is invoked on the
+    train and valid panels every boosting round, per seed.
     """
     if group is None or len(group) == 0:
         return 0.0
     offs = np.concatenate([[0], np.cumsum(group.astype(int))])
-    vals = []
+    gid = np.repeat(np.arange(len(group)), group.astype(int))
-    for i in range(len(group)):
+    rp = _group_averaged_rank(preds, gid, offs)
-        s = slice(offs[i], offs[i + 1])
+    rl = _group_averaged_rank(labels, gid, offs)
-        p, l = preds[s], labels[s]
+    n_g = group.astype(float)
-        if len(p) < 3 or np.std(p) == 0 or np.std(l) == 0:
+    s_p = np.bincount(gid, weights=rp)
-            continue
+    s_l = np.bincount(gid, weights=rl)
-        vals.append(np.corrcoef(pd.Series(p).rank(), pd.Series(l).rank())[0, 1])
+    s_pl = np.bincount(gid, weights=rp * rl)
-    return float(np.mean(vals)) if vals else 0.0
+    s_pp = np.bincount(gid, weights=rp * rp)
+    s_ll = np.bincount(gid, weights=rl * rl)
+    cov = n_g * s_pl - s_p * s_l
+    var_p = n_g * s_pp - s_p ** 2
+    var_l = n_g * s_ll - s_l ** 2
+    denom = np.sqrt(var_p * var_l)
+    valid = (n_g >= 3) & (denom > 0)
+    corr = np.where(valid, cov / np.where(denom == 0, 1, denom), 0.0)
+    return float(corr[valid].mean()) if valid.any() else 0.0
 
 
 def rankic_feval(preds, dataset):

code/tac-qlib/tac_qlib/contrib/strategy/__init__.py

@@ -1,3 +1,13 @@
+from .kelly_dropout import FractionalKellyDropoutStrategy  # noqa: F401
 from .optimal_stop import OptimalStopControl  # noqa: F401
+from .regime_gate import RegimeGateDropoutStrategy  # noqa: F401
+from .top_bottom import TopBottomDropoutStrategy  # noqa: F401
+from .weekly_rebalance import WeeklyRebalanceDropoutStrategy  # noqa: F401
 
-__all__ = ["OptimalStopControl"]
+__all__ = [
+    "OptimalStopControl",
+    "FractionalKellyDropoutStrategy",
+    "WeeklyRebalanceDropoutStrategy",
+    "TopBottomDropoutStrategy",
+    "RegimeGateDropoutStrategy",
+]

code/tac-qlib/tac_qlib/contrib/strategy/kelly_dropout.py

@@ -0,0 +1,201 @@
+"""Fractional-Kelly dropout strategy for cross-sectional signals.
+
+Sizing rule variant of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy``:
+the topk/n_drop SELECTION is identical to the reference, but the buy size is
+proportional to the score MAGNITUDE (edge) instead of equal-weight, capped at a
+fraction ``cap_frac`` of the equal-weight notional so a single name cannot
+over-concentrate the book.
+
+``cap_frac`` is the fraction of the equal-weight per-name notional that a top
+signal can deploy at most (e.g. 0.5 = at most half the equal-weight size).
+Names whose score is below the median of the buy set get a proportionally
+smaller slice; the residual stays in cash (that is the point of the rule:
+throw away less edge per name, deploy less capital when conviction is low).
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+import numpy as np
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
+
+__all__ = ["FractionalKellyDropoutStrategy"]
+
+DEFAULT_CAP_FRAC = 0.5
+
+
+class FractionalKellyDropoutStrategy(TopkDropoutStrategy):
+    """TopkDropout selection with score-magnitude (fractional-Kelly) sizing.
+
+    Parameters
+    ----------
+    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
+    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
+    cap_frac : max buy notional as a fraction of the equal-weight notional.
+    """
+
+    def __init__(self, *, topk, n_drop, cap_frac: float = DEFAULT_CAP_FRAC, **kwargs):
+        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
+        self.cap_frac = cap_frac
+
+    def generate_trade_decision(self, execute_result=None):
+        import copy
+
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None:
+            return TradeDecisionWO([], self)
+
+        if self.only_tradable:
+
+            def get_first_n(li, n, reverse=False):
+                cur_n = 0
+                res = []
+                for si in reversed(li) if reverse else li:
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    ):
+                        res.append(si)
+                        cur_n += 1
+                        if cur_n >= n:
+                            break
+                return res[::-1] if reverse else res
+
+            def get_last_n(li, n):
+                return get_first_n(li, n, reverse=True)
+
+            def filter_stock(li):
+                return [
+                    si
+                    for si in li
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    )
+                ]
+
+        else:
+
+            def get_first_n(li, n):
+                return list(li)[:n]
+
+            def get_last_n(li, n):
+                return list(li)[-n:]
+
+            def filter_stock(li):
+                return li
+
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        sell_order_list: List[Order] = []
+        buy_order_list: List[Order] = []
+        cash = current_temp.get_cash()
+        current_stock_list = current_temp.get_stock_list()
+        last = pred_score.reindex(current_stock_list).sort_values(ascending=False).index
+
+        if self.method_buy == "top":
+            today = get_first_n(
+                pred_score[~pred_score.index.isin(last)].sort_values(ascending=False).index,
+                self.n_drop + self.topk - len(last),
+            )
+        elif self.method_buy == "random":
+            topk_candi = get_first_n(pred_score.sort_values(ascending=False).index, self.topk)
+            candi = list(filter(lambda x: x not in last, topk_candi))
+            n = self.n_drop + self.topk - len(last)
+            try:
+                today = np.random.choice(candi, n, replace=False)
+            except ValueError:
+                today = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        comb = pred_score.reindex(last.union(pd.Index(today))).sort_values(ascending=False).index
+
+        if self.method_sell == "bottom":
+            sell = last[last.isin(get_last_n(comb, self.n_drop))]
+        elif self.method_sell == "random":
+            candi = filter_stock(last)
+            try:
+                sell = pd.Index(np.random.choice(candi, self.n_drop, replace=False) if len(last) else [])
+            except ValueError:
+                sell = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        buy = today[: len(sell) + self.topk - len(last)]
+        for code in current_stock_list:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.SELL,
+            ):
+                continue
+            if code in sell:
+                time_per_step = self.trade_calendar.get_freq()
+                if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                sell_amount = current_temp.get_stock_amount(code=code)
+                sell_order = Order(
+                    stock_id=code,
+                    amount=sell_amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(sell_order):
+                    sell_order_list.append(sell_order)
+                    trade_val, trade_cost, trade_price = self.trade_exchange.deal_order(
+                        sell_order, position=current_temp
+                    )
+                    cash += trade_val - trade_cost
+
+        if len(buy) == 0:
+            return TradeDecisionWO(sell_order_list, self)
+
+        # ---- fractional-Kelly sizing --------------------------------------
+        # equal-weight notional (reference baseline)
+        eq_notional = cash * self.risk_degree / len(buy)
+        buy_scores = pred_score.reindex(buy).astype(float)
+        lo, hi = buy_scores.min(), buy_scores.max()
+        if hi == lo:
+            w = pd.Series(1.0, index=buy_scores.index)
+        else:
+            w = (buy_scores - lo) / (hi - lo)  # [0,1] edge magnitude
+            w = w.clip(lower=0.0)
+            w_max = w.max()
+            w = w / w_max if w_max > 0 else w  # max == 1.0
+        for code in buy:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.BUY,
+            ):
+                continue
+            buy_price = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=OrderDir.BUY
+            )
+            notional = eq_notional * min(self.cap_frac, float(w.get(code, 0.0)))
+            buy_amount = notional / buy_price
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            buy_amount = self.trade_exchange.round_amount_by_trade_unit(buy_amount, factor)
+            buy_order = Order(
+                stock_id=code,
+                amount=buy_amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            buy_order_list.append(buy_order)
+
+        return TradeDecisionWO(sell_order_list + buy_order_list, self)

code/tac-qlib/tac_qlib/contrib/strategy/regime_gate.py

@@ -0,0 +1,231 @@
+"""HMM-regime overlay TopkDropout strategy.
+
+Regime-gate overlay on ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy``:
+selection and sizing are identical to the reference, but a name is only BOUGHT
+(entry gate) when its per-symbol HMM regime posterior ``sp_hmm_p_regime1`` on
+the signal date is >= ``regime_threshold``; otherwise it is held in cash instead
+of being opened.
+
+The regime posterior is read from the lake feature provider on the fly via
+``qlib.data.D.features`` (field ``$sp_hmm_p_regime1``) for the signal window, so
+no regime column needs to enter the model's ``feature_fields`` — the gate is a
+pure overlay (book ch.01: regime flags regressed as model features, survived
+only as an overlay). The HMM itself was fit with ``fit_end=<train end>`` when
+the lake features were backfilled, so there is no lookahead.
+
+Names already held are NOT force-sold when the regime turns unfavourable
+(entry gate only, matching the queue-10 design).
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+import numpy as np
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
+
+try:
+    from qlib.data import D
+except ImportError:  # pragma: no cover - qlib always present in this stack
+    D = None
+
+__all__ = ["RegimeGateDropoutStrategy"]
+
+DEFAULT_REGIME_THRESHOLD = 0.5
+REGIME_FIELD = "$sp_hmm_p_regime1"
+
+
+class RegimeGateDropoutStrategy(TopkDropoutStrategy):
+    """TopkDropout with an HMM-regime entry gate on buy candidates.
+
+    Parameters
+    ----------
+    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
+    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
+    regime_threshold : minimum ``sp_hmm_p_regime1`` posterior required to open a
+        new position (default 0.5).
+    """
+
+    def __init__(self, *, topk, n_drop, regime_threshold: float = DEFAULT_REGIME_THRESHOLD, **kwargs):
+        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
+        self.regime_threshold = regime_threshold
+
+    def _regime_for(self, codes, pred_start, pred_end) -> pd.Series:
+        """Return {code: sp_hmm_p_regime1} for the signal window (last day)."""
+        if D is None:
+            return pd.Series(dtype=float)
+        try:
+            df = D.features(list(codes), [REGIME_FIELD], start_time=pred_start, end_time=pred_end, freq="day")
+        except Exception:  # noqa: BLE001 - a regime read failure should gate open, not crash
+            return pd.Series(dtype=float)
+        if df is None or len(df) == 0:
+            return pd.Series(dtype=float)
+        # df index is MultiIndex (datetime, instrument); take the last day's values
+        df = df.reset_index()
+        ts_col = "datetime" if "datetime" in df.columns else df.columns[0]
+        sym_col = "instrument" if "instrument" in df.columns else df.columns[1]
+        last_ts = df[ts_col].max()
+        last = df[df[ts_col] == last_ts]
+        out = {}
+        for _, row in last.iterrows():
+            sym = str(row[sym_col]).split("/")[-1].upper()
+            val = row.iloc[-1]
+            out[sym] = float(val) if val == val else np.nan
+        return pd.Series(out)
+
+    def generate_trade_decision(self, execute_result=None):
+        import copy
+
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None:
+            return TradeDecisionWO([], self)
+
+        if self.only_tradable:
+
+            def get_first_n(li, n, reverse=False):
+                cur_n = 0
+                res = []
+                for si in reversed(li) if reverse else li:
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    ):
+                        res.append(si)
+                        cur_n += 1
+                        if cur_n >= n:
+                            break
+                return res[::-1] if reverse else res
+
+            def get_last_n(li, n):
+                return get_first_n(li, n, reverse=True)
+
+            def filter_stock(li):
+                return [
+                    si
+                    for si in li
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    )
+                ]
+
+        else:
+
+            def get_first_n(li, n):
+                return list(li)[:n]
+
+            def get_last_n(li, n):
+                return list(li)[-n:]
+
+            def filter_stock(li):
+                return li
+
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        sell_order_list: List[Order] = []
+        buy_order_list: List[Order] = []
+        cash = current_temp.get_cash()
+        current_stock_list = current_temp.get_stock_list()
+        last = pred_score.reindex(current_stock_list).sort_values(ascending=False).index
+
+        if self.method_buy == "top":
+            today = get_first_n(
+                pred_score[~pred_score.index.isin(last)].sort_values(ascending=False).index,
+                self.n_drop + self.topk - len(last),
+            )
+        elif self.method_buy == "random":
+            topk_candi = get_first_n(pred_score.sort_values(ascending=False).index, self.topk)
+            candi = list(filter(lambda x: x not in last, topk_candi))
+            n = self.n_drop + self.topk - len(last)
+            try:
+                today = np.random.choice(candi, n, replace=False)
+            except ValueError:
+                today = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        comb = pred_score.reindex(last.union(pd.Index(today))).sort_values(ascending=False).index
+
+        if self.method_sell == "bottom":
+            sell = last[last.isin(get_last_n(comb, self.n_drop))]
+        elif self.method_sell == "random":
+            candi = filter_stock(last)
+            try:
+                sell = pd.Index(np.random.choice(candi, self.n_drop, replace=False) if len(last) else [])
+            except ValueError:
+                sell = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        buy = today[: len(sell) + self.topk - len(last)]
+
+        # ---- regime gate -----------------------------------------------------
+        if buy:
+            regime = self._regime_for(buy, pred_start_time, pred_end_time)
+            gated = [c for c in buy if regime.get(c, np.nan) >= self.regime_threshold]
+        else:
+            gated = []
+
+        for code in current_stock_list:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.SELL,
+            ):
+                continue
+            if code in sell:
+                time_per_step = self.trade_calendar.get_freq()
+                if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                sell_amount = current_temp.get_stock_amount(code=code)
+                sell_order = Order(
+                    stock_id=code,
+                    amount=sell_amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(sell_order):
+                    sell_order_list.append(sell_order)
+                    trade_val, trade_cost, trade_price = self.trade_exchange.deal_order(
+                        sell_order, position=current_temp
+                    )
+                    cash += trade_val - trade_cost
+
+        if len(gated) == 0:
+            return TradeDecisionWO(sell_order_list, self)
+
+        value = cash * self.risk_degree / len(gated)
+        for code in gated:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.BUY,
+            ):
+                continue
+            buy_price = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=OrderDir.BUY
+            )
+            buy_amount = value / buy_price
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            buy_amount = self.trade_exchange.round_amount_by_trade_unit(buy_amount, factor)
+            buy_order = Order(
+                stock_id=code,
+                amount=buy_amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            buy_order_list.append(buy_order)
+
+        return TradeDecisionWO(sell_order_list + buy_order_list, self)

code/tac-qlib/tac_qlib/contrib/strategy/top_bottom.py

@@ -0,0 +1,169 @@
+"""Market-neutral top/bottom long-short strategy for cross-sectional signals.
+
+Captures the cross-sectional long-short spread net of costs: buys the top-ranked
+``topk`` names and shorts the bottom-ranked ``topk`` names, equal-weight per
+side, sized to ``risk_degree`` of total value per side. Rebalances daily to the
+current rank (dropout-free: the book converges to the latest top/bottom sets).
+
+The long and short legs use equal notional per side (gross exposure ~2x
+``risk_degree`` of NAV, i.e. approximately market neutral before transaction
+costs). Benchmark neutrality (SPY beta ~ 0) is the secondary sanity metric.
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+import copy
+
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import BaseSignalStrategy
+
+__all__ = ["TopBottomDropoutStrategy"]
+
+DEFAULT_SHORT_LEG = True
+DEFAULT_REBALANCE_DAILY = True
+
+
+class TopBottomDropoutStrategy(BaseSignalStrategy):
+    """Long top-k / short bottom-k equal-weight market-neutral book.
+
+    Parameters
+    ----------
+    topk : number of names on each side (long top-k and short bottom-k).
+    short_leg : whether to open the short side (if False, long-only topk).
+    rebalance_daily : if True rebalance to current rank every day; else keep
+        positions and only refresh on score changes (dropout-style).
+    risk_degree : fraction of total value deployed per side.
+    """
+
+    def __init__(
+        self,
+        *,
+        topk: int = 10,
+        short_leg: bool = DEFAULT_SHORT_LEG,
+        rebalance_daily: bool = DEFAULT_REBALANCE_DAILY,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.topk = topk
+        self.short_leg = short_leg
+        self.rebalance_daily = rebalance_daily
+        self._prev_longs = set()
+        self._prev_shorts = set()
+
+    def generate_trade_decision(self, execute_result=None):
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None or len(pred_score) == 0:
+            return TradeDecisionWO([], self)
+
+        # rank all names; topk longs and topk shorts
+        ranked = pred_score.sort_values(ascending=False)
+        longs = list(ranked.index[: self.topk])
+        shorts = list(ranked.index[-self.topk :]) if self.short_leg else []
+
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        current_codes = set(current_temp.get_stock_list())
+        holdings = {c: current_temp for c in current_codes if abs(current_temp.get_stock_amount(c)) > 1e-6}
+
+        sell_orders: List[Order] = []
+        buy_orders: List[Order] = []
+
+        def _tradable(code, direction):
+            try:
+                return self.trade_exchange.is_stock_tradable(
+                    stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=direction
+                )
+            except TypeError:
+                return self.trade_exchange.is_stock_tradable(
+                    stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+                )
+
+        # determine target set (long/short)
+        target_longs = set(longs)
+        target_shorts = set(shorts)
+
+        # close positions not in the target book
+        for code in list(holdings):
+            if code in target_longs or code in target_shorts:
+                continue
+            amt = abs(current_temp.get_stock_amount(code))
+            o = Order(
+                stock_id=code,
+                amount=amt,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.SELL if code in target_longs else Order.SELL,
+            )
+            if self.trade_exchange.check_order(o):
+                sell_orders.append(o)
+                self.trade_exchange.deal_order(o, position=current_temp)
+
+        # equal-weight notional per side
+        total_value = current_temp.get_cash()
+        for code, pos in holdings.items():
+            if code in target_longs or code in target_shorts:
+                mark = self.trade_exchange.get_deal_price(
+                    stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=Order.SELL
+                )
+                if mark is not None and mark == mark:
+                    total_value += abs(current_temp.get_stock_amount(code)) * mark
+
+        side_notional = total_value * self.risk_degree / max(1, self.topk)
+
+        for code in longs:
+            if code in holdings and abs(current_temp.get_stock_amount(code)) > 1e-6:
+                continue
+            px = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=Order.BUY
+            )
+            if px is None or px != px or px <= 0:
+                continue
+            amount = side_notional / px
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            amount = self.trade_exchange.round_amount_by_trade_unit(amount, factor)
+            o = Order(
+                stock_id=code,
+                amount=amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            if self.trade_exchange.check_order(o):
+                buy_orders.append(o)
+
+        if self.short_leg:
+            for code in shorts:
+                if code in holdings and abs(current_temp.get_stock_amount(code)) > 1e-6:
+                    continue
+                px = self.trade_exchange.get_deal_price(
+                    stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=Order.SELL
+                )
+                if px is None or px != px or px <= 0:
+                    continue
+                amount = side_notional / px
+                factor = self.trade_exchange.get_factor(
+                    stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+                )
+                amount = self.trade_exchange.round_amount_by_trade_unit(amount, factor)
+                o = Order(
+                    stock_id=code,
+                    amount=amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(o):
+                    sell_orders.append(o)
+
+        return TradeDecisionWO(sell_orders + buy_orders, self)

code/tac-qlib/tac_qlib/contrib/strategy/weekly_rebalance.py

@@ -0,0 +1,202 @@
+"""Weekly-rebalance TopkDropout strategy.
+
+Turnover-reduction variant of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy``:
+the topk/n_drop selection and sizing are identical to the reference, but the
+target book is recomputed only on the first trading day of each ISO week; on the
+other days the strategy issues NO orders (holds the book untouched).
+
+The weekly cadence is derived from the qlib trade calendar: a rebalance happens
+when the current trade step's date belongs to a different ISO ``(year, week)``
+than the previous trade step. ``hold_band_pct`` (default 0) optionally skips
+tiny rebalances: when a name's existing position differs from the new target by
+less than this fraction, no order is generated for it.
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+import numpy as np
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
+
+__all__ = ["WeeklyRebalanceDropoutStrategy"]
+
+DEFAULT_HOLD_BAND_PCT = 0.0
+
+
+class WeeklyRebalanceDropoutStrategy(TopkDropoutStrategy):
+    """TopkDropout rebalanced once per ISO week; holds otherwise.
+
+    Parameters
+    ----------
+    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
+    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
+    hold_band_pct : skip order for a name whose deviation from target weight is
+        below this fraction of the target (no-trade buffer band).
+    """
+
+    def __init__(self, *, topk, n_drop, hold_band_pct: float = DEFAULT_HOLD_BAND_PCT, **kwargs):
+        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
+        self.hold_band_pct = hold_band_pct
+
+    @staticmethod
+    def _iso_week(ts) -> tuple:
+        return (ts.year, ts.week)
+
+    def generate_trade_decision(self, execute_result=None):
+        import copy
+
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+
+        cur_week = self._iso_week(trade_start_time)
+        prev_week = getattr(self, "_last_week", None)
+        self._last_week = cur_week
+
+        if prev_week is not None and prev_week == cur_week:
+            # not the first trading day of this ISO week -> hold
+            return TradeDecisionWO([], self)
+
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None:
+            return TradeDecisionWO([], self)
+
+        if self.only_tradable:
+
+            def get_first_n(li, n, reverse=False):
+                cur_n = 0
+                res = []
+                for si in reversed(li) if reverse else li:
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    ):
+                        res.append(si)
+                        cur_n += 1
+                        if cur_n >= n:
+                            break
+                return res[::-1] if reverse else res
+
+            def get_last_n(li, n):
+                return get_first_n(li, n, reverse=True)
+
+            def filter_stock(li):
+                return [
+                    si
+                    for si in li
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    )
+                ]
+
+        else:
+
+            def get_first_n(li, n):
+                return list(li)[:n]
+
+            def get_last_n(li, n):
+                return list(li)[-n:]
+
+            def filter_stock(li):
+                return li
+
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        sell_order_list: List[Order] = []
+        buy_order_list: List[Order] = []
+        cash = current_temp.get_cash()
+        current_stock_list = current_temp.get_stock_list()
+        last = pred_score.reindex(current_stock_list).sort_values(ascending=False).index
+
+        if self.method_buy == "top":
+            today = get_first_n(
+                pred_score[~pred_score.index.isin(last)].sort_values(ascending=False).index,
+                self.n_drop + self.topk - len(last),
+            )
+        elif self.method_buy == "random":
+            topk_candi = get_first_n(pred_score.sort_values(ascending=False).index, self.topk)
+            candi = list(filter(lambda x: x not in last, topk_candi))
+            n = self.n_drop + self.topk - len(last)
+            try:
+                today = np.random.choice(candi, n, replace=False)
+            except ValueError:
+                today = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        comb = pred_score.reindex(last.union(pd.Index(today))).sort_values(ascending=False).index
+
+        if self.method_sell == "bottom":
+            sell = last[last.isin(get_last_n(comb, self.n_drop))]
+        elif self.method_sell == "random":
+            candi = filter_stock(last)
+            try:
+                sell = pd.Index(np.random.choice(candi, self.n_drop, replace=False) if len(last) else [])
+            except ValueError:
+                sell = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        buy = today[: len(sell) + self.topk - len(last)]
+        for code in current_stock_list:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.SELL,
+            ):
+                continue
+            if code in sell:
+                time_per_step = self.trade_calendar.get_freq()
+                if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                sell_amount = current_temp.get_stock_amount(code=code)
+                sell_order = Order(
+                    stock_id=code,
+                    amount=sell_amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(sell_order):
+                    sell_order_list.append(sell_order)
+                    trade_val, trade_cost, trade_price = self.trade_exchange.deal_order(
+                        sell_order, position=current_temp
+                    )
+                    cash += trade_val - trade_cost
+
+        if len(buy) == 0:
+            return TradeDecisionWO(sell_order_list, self)
+
+        value = cash * self.risk_degree / len(buy)
+        for code in buy:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.BUY,
+            ):
+                continue
+            buy_price = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=OrderDir.BUY
+            )
+            buy_amount = value / buy_price
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            buy_amount = self.trade_exchange.round_amount_by_trade_unit(buy_amount, factor)
+            buy_order = Order(
+                stock_id=code,
+                amount=buy_amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            buy_order_list.append(buy_order)
+
+        return TradeDecisionWO(sell_order_list + buy_order_list, self)

code/tac-qlib/tac_qlib/data/config.py

@@ -6,10 +6,11 @@ The lake is a hive-partitioned parquet store (see ``tac-engine/skills/tradeac-la
     ├── market=US/
     │   └── timeframe=1d/
     │       └── symbol=AAPL.parquet        # OHLCV bars: t, date, o, h, l, c, v, n, vw
-    ├── features/                          # ta-lib indicators, wide format
+    ├── features/                          # indicators, wide format, family tier
     │   └── market=US/
     │       └── timeframe=1d/
-    │           └── symbol=AAPL.parquet    # t, sma_5, sma_20, rsi_14, ...
+    │           ├── family=ta/symbol=AAPL.parquet  # t, sma_5, sma_20, rsi_14, ...
+    │           └── family=sp/symbol=AAPL.parquet  # t, sp_ou_*, sp_hmm_*, ...
     ├── calendar.parquet                   # trading days per market
     ├── coverage.parquet                   # per (market,timeframe,symbol) loaded windows
     └── symbols.parquet                    # asset master
@@ -107,8 +108,34 @@ class LakeConfig:
         return self.lake_root / "features" / f"market={self.market}" / f"timeframe={timeframe}"
 
     def features_path(self, timeframe: str, symbol: str) -> Path:
+        # Legacy flat path (no family tier). Prefer `load_features` which
+        # resolves the family=ta|sp partition layout.
         return self.features_dir(timeframe) / f"symbol={str(symbol).upper()}.parquet"
 
+    def load_features(self, timeframe: str, symbol: str) -> pd.DataFrame:
+        """All feature columns for a symbol, merging the `family=ta` and
+        `family=sp` partitions by timestamp. Returns an empty frame when no
+        feature files exist (legacy flat layout falls back transparently)."""
+        sym = str(symbol).upper()
+        frames = []
+        for family in ("ta", "sp"):
+            p = self.features_dir(timeframe) / f"family={family}" / f"symbol={sym}.parquet"
+            if p.exists():
+                frames.append(pd.read_parquet(p))
+        if not frames:
+            flat = self.features_dir(timeframe) / f"symbol={sym}.parquet"
+            if flat.exists():
+                return pd.read_parquet(flat)
+            return pd.DataFrame()
+        if len(frames) == 1:
+            return frames[0]
+        merged = frames[0]
+        for extra in frames[1:]:
+            merged = merged.merge(extra, on="t", how="outer", suffixes=("", "_dup"))
+            for c in [c for c in merged.columns if c.endswith("_dup")]:
+                merged = merged.drop(columns=c)
+        return merged
+
     def calendar_path(self) -> Path:
         return self.lake_root / "calendar.parquet"
 

code/tac-qlib/tac_qlib/data/providers.py

@@ -173,8 +173,7 @@ class LakeFeatureProvider(FeatureProvider):
     def _load_feature_df(self, instrument: str, timeframe: str) -> pd.DataFrame:
         key = (instrument, timeframe)
         if key not in self._feature_cache:
-            p = self.cfg.features_path(timeframe, instrument)
+            self._feature_cache[key] = self.cfg.load_features(timeframe, instrument)
-            self._feature_cache[key] = pd.read_parquet(p) if p.exists() else pd.DataFrame()
         return self._feature_cache[key]
 
     @staticmethod

workflows/ablate_baseline_all_sp_fields.yaml

@@ -1,133 +0,0 @@
-# -----------------------------------------------------------------------------
-# ABLATION A (baseline): LightGBM with RankIC early-stopping on the 50-ETF SP-5d
-# panel, using ALL 24 sp_* feature columns (ou,hmm,jump,har,trend,hurst,
-# signature). Copy of the canonical workflow_lgb_sp5d_rankic.yaml with a
-# distinct experiment name so the ablation runs are isolated.
-#
-# Run:
-#   rd_run_workflow config_path=tac-qlib/workflows/ablate_baseline_all_sp_fields.yaml \
-#       experiment_name=tac-rd-rank-ablate
-# -----------------------------------------------------------------------------
-{%- set LAKE = TAC_LAKE_DIR %}
-{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-            markets: {}
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-rank-ablate"
-
-task:
-    model:
-        class: RankICLGBModel
-        module_path: tac_qlib.contrib.model.rank_gbdt
-        kwargs:
-            loss: mse
-            learning_rate: 0.02
-            num_leaves: 31
-            n_estimators: 3000
-            num_boost_round: 3000
-            early_stopping_rounds: 200
-            min_data_in_leaf: 20
-            lambda_l2: 0.5
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.1
-            reg_lambda: 1.0
-            seed: 42
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: "{{ UNIVERSE }}"
-                    start_time: 2015-01-03
-                    end_time: 2026-08-10
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2025-09-01
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
-                    infer_processors:
-                        - class: DropAllNaN
-                          kwargs: {}
-                        - class: ProcessInf
-                          kwargs: {}
-                        - class: CSRankNorm
-                          kwargs: {}
-                        - class: ZScoreNorm
-                          kwargs: {}
-                        - class: Fillna
-                          kwargs: {}
-            segments:
-                train: [2015-01-03, 2025-09-01]
-                valid: [2025-09-03, 2026-01-03]
-                test: [2026-01-04, 2026-08-10]
-
-    record:
-        - class: SignalRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: {}
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              ana_long_short: true
-              ann_scaler: 252
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: TopkDropoutStrategy
-                      module_path: qlib.contrib.strategy
-                      kwargs:
-                          signal: "<PRED>"
-                          topk: 10
-                          n_drop: 2
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2026-01-04
-                      end_time: 2026-08-10
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: "{{ UNIVERSE }}"
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d

workflows/ablate_generic_only_sp_fields.yaml

@@ -1,134 +0,0 @@
-# -----------------------------------------------------------------------------
-# ABLATION B (generic-only): same panel/model as the baseline, but feature
-# fields restricted to the model-free / generic stochastic-process families
-# (jump,har,trend,hurst,signature). Drops the model-specific ou (OU/AR-1
-# half-life) and hmm (2-state regime) families to test whether the generic
-# families alone dominate the rank dimension.
-#
-# Run:
-#   rd_run_workflow config_path=tac-qlib/workflows/ablate_generic_only_sp_fields.yaml \
-#       experiment_name=tac-rd-rank-ablate
-# -----------------------------------------------------------------------------
-{%- set LAKE = TAC_LAKE_DIR %}
-{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-            markets: {}
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-rank-ablate"
-
-task:
-    model:
-        class: RankICLGBModel
-        module_path: tac_qlib.contrib.model.rank_gbdt
-        kwargs:
-            loss: mse
-            learning_rate: 0.02
-            num_leaves: 31
-            n_estimators: 3000
-            num_boost_round: 3000
-            early_stopping_rounds: 200
-            min_data_in_leaf: 20
-            lambda_l2: 0.5
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.1
-            reg_lambda: 1.0
-            seed: 42
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: "{{ UNIVERSE }}"
-                    start_time: 2015-01-03
-                    end_time: 2026-08-10
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2025-09-01
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
-                    infer_processors:
-                        - class: DropAllNaN
-                          kwargs: {}
-                        - class: ProcessInf
-                          kwargs: {}
-                        - class: CSRankNorm
-                          kwargs: {}
-                        - class: ZScoreNorm
-                          kwargs: {}
-                        - class: Fillna
-                          kwargs: {}
-            segments:
-                train: [2015-01-03, 2025-09-01]
-                valid: [2025-09-03, 2026-01-03]
-                test: [2026-01-04, 2026-08-10]
-
-    record:
-        - class: SignalRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: {}
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              ana_long_short: true
-              ann_scaler: 252
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: TopkDropoutStrategy
-                      module_path: qlib.contrib.strategy
-                      kwargs:
-                          signal: "<PRED>"
-                          topk: 10
-                          n_drop: 2
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2026-01-04
-                      end_time: 2026-08-10
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: "{{ UNIVERSE }}"
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d

workflows/exp12_isolation_ensemble.yaml

@@ -1,141 +0,0 @@
-# -----------------------------------------------------------------------------
-# ISOLATION: multi-seed RankIC ensemble, ablate-B generic-only feature set.
-#
-# Isolates the ensemble effect on the SP-5d rank signal. Same panel, segments,
-# history (full backfilled 2016+) and feature set as the exp-9 ablate-B winner
-# (generic-only sp_* families: jump,har,trend,hurst,signature), but replaces the
-# single RankICLGBModel with a 5-seed RankICEnsembleLGBModel (42,7,2026,99,123)
-# that averages per-day predictions.
-#
-# Differs from exp-15 (tac-rd-rank-ensemble, mlflow exp 15) ONLY by dropping the
-# TA subset (rsi_14,roc_10,macd_hist,willr_14,atr_14) and the inter-asset xr_*
-# features, so any change vs exp-15 is attributable to the feature set alone,
-# and any change vs exp-9 is attributable to the ensemble + full history alone.
-#
-# Run:
-#   rd_run_workflow config_path=experiments/workflows/exp12_isolation_ensemble.yaml \
-#       experiment_name=tac-rd-rank-ensemble-isolated
-# -----------------------------------------------------------------------------
-{%- set LAKE = TAC_LAKE_DIR %}
-{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-            markets: {}
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs:
-            lake_root: "{{ LAKE }}"
-            market: US
-
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs:
-            uri: "sqlite:///mlruns.db"
-            default_exp_name: "tac-rd-rank-ensemble-isolated"
-
-task:
-    model:
-        class: RankICEnsembleLGBModel
-        module_path: tac_qlib.contrib.model.rank_ensemble
-        kwargs:
-            loss: mse
-            learning_rate: 0.02
-            num_leaves: 31
-            n_estimators: 3000
-            num_boost_round: 3000
-            early_stopping_rounds: 200
-            min_data_in_leaf: 20
-            lambda_l2: 0.5
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.1
-            reg_lambda: 1.0
-            seeds: "42,7,2026,99,123"
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: "{{ UNIVERSE }}"
-                    start_time: 2015-01-03
-                    end_time: 2026-08-14
-                    fit_start_time: 2016-01-04
-                    fit_end_time: 2025-09-01
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
-                    infer_processors:
-                        - class: DropAllNaN
-                          kwargs: {}
-                        - class: ProcessInf
-                          kwargs: {}
-                        - class: CSRankNorm
-                          kwargs: {}
-                        - class: ZScoreNorm
-                          kwargs: {}
-                        - class: Fillna
-                          kwargs: {}
-            segments:
-                train: [2016-01-04, 2025-09-01]
-                valid: [2025-09-03, 2026-01-03]
-                test: [2026-01-04, 2026-08-10]
-
-    record:
-        - class: SignalRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: {}
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              ana_long_short: true
-              ann_scaler: 252
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: TopkDropoutStrategy
-                      module_path: qlib.contrib.strategy
-                      kwargs:
-                          signal: "<PRED>"
-                          topk: 10
-                          n_drop: 2
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2026-01-04
-                      end_time: 2026-08-10
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: "{{ UNIVERSE }}"
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d

workflows/exp17-short-lake/workflow.yaml

@@ -1,96 +0,0 @@
-# -----------------------------------------------------------------------------
-# exp17 — short 3-month lake workflow (trace id 17)
-# train 2026-05-01..06-30 / valid 07-01..07-15 / test 07-16..08-14
-# 20-ETF universe, 5d forward label, LGBModel mse, TopkDropout topk=3/n_drop=1
-# -----------------------------------------------------------------------------
-{%- set LAKE = TAC_LAKE_DIR %}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-short-lake" }
-
-task:
-    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
-        kwargs:
-            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: SPY,QQQ,IWM,EEM,TLT,GLD,USO,SMH,XLF,XLE,XLY,XLK,XLV,XBI,IBB,KWEB,FXI,EWJ,VNQ,GDX
-                    start_time: 2026-05-01
-                    end_time: 2026-08-14
-                    fit_start_time: 2026-05-01
-                    fit_end_time: 2026-06-30
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-            segments:
-                train: [2026-05-01, 2026-06-30]
-                valid: [2026-07-01, 2026-07-15]
-                test:  [2026-07-16, 2026-08-14]
-
-    record:
-        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: TopkDropoutStrategy
-                      module_path: qlib.contrib.strategy
-                      kwargs:
-                          signal: "<PRED>"
-                          topk: 3
-                          n_drop: 1
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2026-07-16
-                      end_time: 2026-08-14
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: SPY,QQQ,IWM,EEM,TLT,GLD,USO,SMH,XLF,XLE,XLY,XLK,XLV,XBI,IBB,KWEB,FXI,EWJ,VNQ,GDX
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d