finish experiment 21 (exp/21-clean-lake-re-execution-of-the-tac-rd-ra)

code/MANIFEST.txt

@@ -0,0 +1,27 @@
+# TradeAC custom-qlib-code snapshot (auto-generated)
+# parent repo HEAD : ab919245c2f3d6881cd6e16e77e583bbb6d5b000
+#   tac-qlib/tac_qlib/contrib
+#   tac-qlib/tac_qlib/data
+# per-file hashes (git hash-object):
+  1b6298c4a5652f2e863cbdc385a1014a570fcd59  tac-qlib/tac_qlib/contrib/__init__.py
+  bb903ca28c70ae5802d673d5b85481679f58c286  tac-qlib/tac_qlib/contrib/__pycache__/__init__.cpython-312.pyc
+  c76a9f17f680e74eea766eff27f7624359749ed6  tac-qlib/tac_qlib/contrib/data/__init__.py
+  9749bb730880371ea7bf9bf0c5ffd78fcf6b5a91  tac-qlib/tac_qlib/contrib/data/__pycache__/__init__.cpython-312.pyc
+  9eb94cfac5d41ae20acb12612c219f210d463ab4  tac-qlib/tac_qlib/contrib/data/__pycache__/handler.cpython-312.pyc
+  871ff1e163c29261f140c3f53d42a41e6504c779  tac-qlib/tac_qlib/contrib/data/handler.py
+  b151d139a0dcde87d74b21e7c4b729176ba5c39b  tac-qlib/tac_qlib/contrib/model/__init__.py
+  d209c3e3e8c2a8683cd337ff3b10c04015f16dc6  tac-qlib/tac_qlib/contrib/model/__pycache__/__init__.cpython-312.pyc
+  532527ebe81b269f723c806286122fb5483d0379  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_ensemble.cpython-312.pyc
+  1681c9bc021188c0f134e33e6402f521a9d46e9c  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
+  4afcf9058231111c412925f4c4b84e81d656db87  tac-qlib/tac_qlib/contrib/strategy/__init__.py
+  a6fb3d6c2d111b7ad423939582df67aacb36fead  tac-qlib/tac_qlib/contrib/strategy/__pycache__/__init__.cpython-312.pyc
+  44ed28758151eb7fa4d388646cb1c6f04b450d8c  tac-qlib/tac_qlib/contrib/strategy/__pycache__/optimal_stop.cpython-312.pyc
+  79aaad9e39fcc740a773f4f63c512ce1086cfde0  tac-qlib/tac_qlib/contrib/strategy/optimal_stop.py
+  92e6e90eb0cd0a25142034560f27adb6b705b1a8  tac-qlib/tac_qlib/data/__init__.py
+  319e3f728b093d6c483eb29de42617a45ee88830  tac-qlib/tac_qlib/data/__pycache__/__init__.cpython-312.pyc
+  08a7dcbcf3f34bdb784d3b16c04daf265d04ae5f  tac-qlib/tac_qlib/data/__pycache__/config.cpython-312.pyc
+  47337bd1e54b6e333f26a9088d645ed48fbc44f6  tac-qlib/tac_qlib/data/__pycache__/providers.cpython-312.pyc
+  686d36f6d101c547491ca866aa143aa542e17518  tac-qlib/tac_qlib/data/config.py
+  d9f839be30026f337754a3f015425a8efdbe8e2a  tac-qlib/tac_qlib/data/providers.py

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

@@ -0,0 +1,11 @@
+from . import data  # noqa: F401  (registers tac_qlib.contrib.data)
+from . import model, strategy  # noqa: F401
+from .data import TACHandler  # noqa: F401
+from .model import RankICLGBModel  # noqa: F401
+from .strategy import OptimalStopControl  # noqa: F401
+
+__all__ = [
+    "TACHandler",
+    "RankICLGBModel",
+    "OptimalStopControl",
+]

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

@@ -0,0 +1,3 @@
+from .handler import TACHandler
+
+__all__ = ["TACHandler"]

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

@@ -0,0 +1,236 @@
+"""TACHandler: a qlib DataHandlerLP that builds datasets from the TradeAC lake.
+
+This is the "custom DataHandler" entry point (Option B): the handler is referenced from the
+workflow yaml's ``dataset.handler`` and reads OHLCV + pre-computed ta-lib features straight
+from the lake parquet files through ``QLibDataLoader`` + the tac_qlib feature provider.
+
+The standard qlib processor pipeline (``infer_processors`` / ``learn_processors``) still runs
+on top, so existing recipes such as ``DropnaLabel``, ``CSZScoreNorm`` or ``RobustZScoreNorm``
+keep working unchanged.
+"""
+
+from __future__ import annotations
+
+import os
+from inspect import getfullargspec
+from typing import List, Optional, Tuple, Union
+
+from qlib.data.dataset import processor as processor_module
+from qlib.data.dataset.handler import DataHandlerLP
+from qlib.utils import get_callable_kwargs
+
+from ...data.config import (
+    LakeConfig,
+    timeframe_for_freq,
+    NON_FEATURE_COLUMNS,
+)
+
+DEFAULT_INFER_PROCESSORS = [
+    {"class": "DropAllNaN", "kwargs": {}},
+    {"class": "ProcessInf", "kwargs": {}},
+    {"class": "ZScoreNorm", "kwargs": {}},
+    {"class": "Fillna", "kwargs": {}},
+]
+DEFAULT_LEARN_PROCESSORS = [
+    {"class": "DropnaLabel"},
+    {"class": "CSZScoreNorm", "kwargs": {"fields_group": "label"}},
+]
+
+#: always include raw OHLCV; ta-lib columns are discovered from the lake and appended.
+RAW_FEATURE_FIELDS = ("$open", "$high", "$low", "$close", "$vwap", "$volume")
+
+DEFAULT_LABEL = "Ref($close,-2)/Ref($close,-1)-1"
+
+
+def check_transform_proc(proc_l, fit_start_time, fit_end_time):
+    """Port of ``qlib.contrib.data.handler.check_transform_proc`` (inject fit window into procs)."""
+    new_l = []
+    for p in proc_l:
+        if not isinstance(p, processor_module.Processor):
+            klass, pkwargs = get_callable_kwargs(p, processor_module)
+            args = getfullargspec(klass).args
+            if "fit_start_time" in args and "fit_end_time" in args:
+                assert fit_start_time is not None and fit_end_time is not None, (
+                    "Make sure `fit_start_time` and `fit_end_time` are not None."
+                )
+                pkwargs.update({"fit_start_time": fit_start_time, "fit_end_time": fit_end_time})
+            proc_config = {"class": klass.__name__, "kwargs": pkwargs}
+            if isinstance(p, dict) and "module_path" in p:
+                proc_config["module_path"] = p["module_path"]
+            new_l.append(proc_config)
+        else:
+            new_l.append(p)
+    return new_l
+
+
+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.
+
+    Returns sorted field names (without the ``$`` prefix). Empty if no features are persisted.
+    """
+    cfg = LakeConfig(lake_root, market)
+    feat_dir = cfg.features_dir(timeframe)
+    if not feat_dir.exists():
+        return []
+    import pyarrow.parquet as pq
+
+    common = None
+    for p in sorted(feat_dir.glob("symbol=*.parquet")):
+        try:
+            cols = set(pq.read_schema(p).names) - set(NON_FEATURE_COLUMNS)
+        except Exception:  # pragma: no cover - skip unreadable files
+            continue
+        common = cols if common is None else (common & cols)
+        if not common:
+            break
+    return sorted(common) if common else []
+
+
+class DropAllNaN(processor_module.Processor):
+    """Drop feature columns that are all-NaN over the fit window.
+
+    The lake can hold fully-empty indicator columns (e.g. a ta-lib output that was NaN
+    from the start). Such columns carry no learnable signal and make ``ZScoreNorm.fit``
+    warn on empty slices, so we drop them before any other processor runs. The drop set
+    is fixed on the fit window once (during ``fit``), then applied consistently to every
+    segment so train/valid/test keep identical feature columns.
+    """
+
+    def __init__(self, fit_start_time=None, fit_end_time=None):
+        self.fit_start_time = fit_start_time
+        self.fit_end_time = fit_end_time
+        self.cols_to_drop = []
+
+    def fit(self, df=None):
+        if df is None or len(df) == 0:
+            return self
+        window = df
+        if self.fit_start_time is not None and self.fit_end_time is not None:
+            try:
+                from qlib.data.dataset.utils import fetch_df_by_index
+
+                window = fetch_df_by_index(
+                    df, slice(self.fit_start_time, self.fit_end_time), level="datetime"
+                )
+            except Exception:  # pragma: no cover - defensive
+                window = df
+        if len(window) == 0:
+            return self
+        self.cols_to_drop = [c for c in window.columns if window[c].isna().all()]
+        return self
+
+    def __call__(self, df):
+        if self.cols_to_drop:
+            return df.drop(columns=self.cols_to_drop, errors="ignore")
+        return df
+
+
+class TACHandler(DataHandlerLP):
+    """DataHandlerLP backed by the TradeAC parquet lake.
+
+    Parameters mirror ``Alpha158``: ``instruments``/``start_time``/``end_time``/``freq`` define
+    the queried window; ``feature_fields`` selects the features (default: raw OHLCV + all common
+    ta-lib columns found in the lake); ``label`` is a qlib expression for the target.
+    """
+
+    def __init__(
+        self,
+        instruments="all",
+        start_time=None,
+        end_time=None,
+        freq="day",
+        infer_processors=DEFAULT_INFER_PROCESSORS,
+        learn_processors=DEFAULT_LEARN_PROCESSORS,
+        fit_start_time=None,
+        fit_end_time=None,
+        process_type=DataHandlerLP.PTYPE_A,
+        filter_pipe=None,
+        feature_fields=None,
+        label=DEFAULT_LABEL,
+        lake_root=None,
+        market="US",
+        **kwargs,
+    ):
+        # default the processor fit window to the queried window (like Alpha158 without a split)
+        if fit_start_time is None:
+            fit_start_time = start_time
+        if fit_end_time is None:
+            fit_end_time = end_time
+
+        infer_processors = check_transform_proc(infer_processors, fit_start_time, fit_end_time)
+        learn_processors = check_transform_proc(learn_processors, fit_start_time, fit_end_time)
+
+        feature_fields = self._normalize_feature_fields(feature_fields, freq, lake_root, market)
+        if not feature_fields:
+            raise ValueError(
+                "no feature fields available for the lake; set `feature_fields` explicitly "
+                "(e.g. ['$close', '$rsi_14', '$sma_20'])"
+            )
+
+        label_expr, label_names = self._normalize_label(label)
+
+        data_loader = {
+            "class": "QlibDataLoader",
+            "kwargs": {
+                "config": {
+                    "feature": (feature_fields, feature_fields),
+                    "label": (label_expr, label_names),
+                },
+                "filter_pipe": filter_pipe,
+                "freq": freq,
+            },
+        }
+        super().__init__(
+            instruments=instruments,
+            start_time=start_time,
+            end_time=end_time,
+            data_loader=data_loader,
+            infer_processors=infer_processors,
+            learn_processors=learn_processors,
+            process_type=process_type,
+            **kwargs,
+        )
+
+    # ------------------------------------------------------------------ config
+    @staticmethod
+    def _normalize_feature_fields(feature_fields, freq, lake_root, market) -> List[str]:
+        if feature_fields is None:
+            common = get_common_feature_fields(lake_root, market, timeframe_for_freq(freq))
+            feature_fields = list(RAW_FEATURE_FIELDS) + ["$" + f for f in common if "$" + f not in RAW_FEATURE_FIELDS]
+        elif isinstance(feature_fields, str):
+            feature_fields = [f.strip() for f in feature_fields.split(",") if f.strip()]
+        fields = [f if f.startswith("$") else "$" + f for f in feature_fields]
+        # de-dup while preserving order
+        seen, out = set(), []
+        for f in fields:
+            if f not in seen:
+                seen.add(f)
+                out.append(f)
+        return out
+
+    @staticmethod
+    def _normalize_label(label) -> Tuple[List[str], List[str]]:
+        if isinstance(label, str):
+            return [label], ["LABEL0"]
+        if isinstance(label, (list, tuple)):
+            if len(label) == 2 and isinstance(label[0], str):
+                return [label[0]], list(label[1]) if isinstance(label[1], (list, tuple)) else [label[1]]
+            return list(label), ["LABEL%d" % i for i in range(len(label))]
+        raise TypeError(f"unsupported label config: {label!r}")
+
+    # ------------------------------------------------------------------ utils
+    def get_label_config(self):
+        return DEFAULT_LABEL
+
+    @staticmethod
+    def discover_feature_fields(lake_root=None, market="US", freq="day") -> List[str]:
+        return get_common_feature_fields(lake_root, market, timeframe_for_freq(freq))
+
+
+__all__ = ["TACHandler", "DropAllNaN", "get_common_feature_fields"]
+
+
+# Make `DropAllNaN` resolvable by bare name from processor configs (e.g. the default
+# ``infer_processors`` and workflow yamls that reference it without a ``module_path``),
+# mirroring how qlib registers its own processors in ``qlib.data.dataset.processor``.
+processor_module.DropAllNaN = DropAllNaN

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

@@ -0,0 +1,4 @@
+from .rank_ensemble import RankICEnsembleLGBModel  # noqa: F401
+from .rank_gbdt import RankICLGBModel, rankic_feval  # noqa: F401
+
+__all__ = ["RankICLGBModel", "rankic_feval", "RankICEnsembleLGBModel"]

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

@@ -0,0 +1,189 @@
+"""Seed-ensembled LightGBM that early-stops on cross-sectional RankIC.
+
+``RankICEnsembleLGBModel`` wraps ``RankICLGBModel`` (per-day RankIC feval +
+``metric='None'`` + ``first_metric_only`` early stopping) over a seed ensemble:
+one sub-model is trained per seed with identical hyper-parameters, and
+predictions are averaged across seeds. This is the model class the
+``tac-rd-rank-ensemble-isolated`` reference run wires into its workflow
+(``module_path: tac_qlib.contrib.model.rank_ensemble``).
+
+The ensemble inherits the RankIC early-stopping behaviour of the single-seed
+model (valid RankIC drives the stopping iteration) while the seed averaging
+stabilizes the prediction against any single seed's early-stopping path.
+
+Training is parallelized: the seed sub-models train in a thread pool —
+``lgb.train`` is C++ and releases the GIL, so concurrent seeds do not block on
+the GIL (5 seeds ~40min/5 on this box). Measured on a 6-physical-core / 12 SMT
+host: the seeds scale ~2x, not linearly — the runs are memory-bandwidth bound
+and each Booster caps its threads at ``cores // workers`` so 5 concurrent
+boosters don't oversubscribe; larger-core hosts scale better. The qlib data
+pipeline is warmed once on the calling thread (fills the handler cache), and
+each worker then prepares its **own** ``lgb.Dataset`` (independent handle, so
+no concurrent ``construct()`` on a shared handle — LightGBM's ``Dataset`` is
+not thread-safe to build). qlib's ``R`` recorder is also not thread-safe, so
+the per-seed evaluation curves are logged on the calling thread after the pool
+finishes.
+
+Wired into a workflow yaml like:
+
+    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"
+            parallel: 5
+
+Any ``**kwargs`` other than ``seeds``/``parallel`` are forwarded unchanged to
+every ``RankICLGBModel`` sub-model (same params, different ``seed``).
+"""
+
+from __future__ import annotations
+
+import os
+from concurrent.futures import ThreadPoolExecutor
+from typing import List, Optional
+
+import pandas as pd
+
+from qlib.data.dataset import DatasetH
+from qlib.data.dataset.handler import DataHandlerLP
+
+from tac_qlib.contrib.model.rank_gbdt import RankICLGBModel
+
+__all__ = ["RankICEnsembleLGBModel"]
+
+
+class RankICEnsembleLGBModel(RankICLGBModel):
+    """Seed ensemble of RankIC-early-stopping LightGBM models.
+
+    Parameters
+    ----------
+    seeds : comma-separated integers, one sub-model per seed.
+    parallel : number of seeds to train concurrently. ``0`` (default) = auto
+        (all seeds, bounded by the available cores); ``1`` = sequential.
+    **kwargs : forwarded to every ``RankICLGBModel`` sub-model (model
+        hyper-parameters). ``seeds``/``parallel`` are consumed here and not
+        forwarded.
+    """
+
+    def __init__(self, seeds: str = "42", parallel: int = 0, **kwargs):
+        self.seeds = [int(s.strip()) for s in str(seeds).split(",") if s.strip()]
+        if not self.seeds:
+            raise ValueError("seeds must contain at least one integer")
+        self.parallel = int(parallel)
+        # drop seed/parallel handling from the base kwargs, keep everything else
+        self._model_kwargs = dict(kwargs)
+        super().__init__(**self._model_kwargs)
+        self._models: List[RankICLGBModel] = []
+
+    # --------------------------------------------------------------- helpers
+    @staticmethod
+    def _cores() -> int:
+        try:
+            return max(1, len(os.sched_getaffinity(0)))
+        except AttributeError:
+            return max(1, os.cpu_count() or 1)
+
+    def _worker_count(self) -> int:
+        if self.parallel > 0:
+            return min(len(self.seeds), self.parallel)
+        return min(len(self.seeds), self._cores())
+
+    # ------------------------------------------------------------------ fit
+    def fit(
+        self,
+        dataset: DatasetH,
+        num_boost_round: Optional[int] = None,
+        early_stopping_rounds: Optional[int] = None,
+        verbose_eval: int = 20,
+        evals_result=None,
+        reweighter=None,
+        **kwargs,
+    ):
+        """Train one RankICLGBModel per seed and keep them for prediction.
+
+        The qlib data pipeline is warmed once on this thread (handler cache),
+        then each seed sub-model trains in a parallel worker thread on its own
+        ``lgb.Dataset`` (LightGBM releases the GIL in ``lgb.train``). Evals
+        are logged on this thread after the pool (qlib's ``R`` is not
+        thread-safe).
+        """
+        n_round = num_boost_round or self.num_boost_round
+        n_es = early_stopping_rounds or self.early_stopping_rounds
+
+        if len(self.seeds) == 1:
+            m = RankICLGBModel(seed=self.seeds[0], **self._model_kwargs)
+            m.fit(
+                dataset,
+                num_boost_round=n_round,
+                early_stopping_rounds=n_es,
+                verbose_eval=verbose_eval,
+                evals_result=evals_result,
+                reweighter=reweighter,
+                **kwargs,
+            )
+            self._models = [m]
+            return
+
+        # Warm the qlib handler cache once on this thread so the workers'
+        # concurrent prepare() calls only hit cached frames (no first-write race).
+        proto = RankICLGBModel(seed=self.seeds[0], **self._model_kwargs)
+        proto._prepare_data(dataset, reweighter)
+
+        workers = self._worker_count()
+        # Cap per-Booster threads so concurrent seeds don't oversubscribe
+        # (LightGBM's num_threads=0 uses ALL cores per Booster).
+        per_booster = max(1, self._cores() // workers)
+
+        def fit_seed(seed):
+            m = RankICLGBModel(seed=seed, **self._model_kwargs)
+            if workers > 1 and "num_threads" not in m.params:
+                m.params["num_threads"] = per_booster
+            ds_l = m._prepare_data(dataset, reweighter)
+            booster, evals, names = m._train_from_datasets(
+                ds_l,
+                num_boost_round=n_round,
+                early_stopping_rounds=n_es,
+                verbose_eval=verbose_eval,
+                **kwargs,
+            )
+            m.model = booster
+            return m, evals, names
+
+        with ThreadPoolExecutor(max_workers=workers) as ex:
+            results = list(ex.map(fit_seed, self.seeds))
+
+        self._models = [m for m, _, _ in results]
+
+        # Merge + log evals on the main thread (qlib's R is not thread-safe).
+        if evals_result is not None:
+            for m, evals, names in results:
+                for k in names:
+                    for key, val in evals.get(k, {}).items():
+                        evals_result.setdefault(f"{k}.seed{m.params['seed']}", {})[key] = val
+        for m, evals, names in results:
+            self._log_evals(evals, names, prefix=f"seed{m.params['seed']}.")
+
+    # -------------------------------------------------------------- predict
+    def predict(self, dataset: DatasetH, segment="test") -> pd.Series:
+        """Average the per-seed predictions over the given segment."""
+        if not self._models:
+            raise ValueError("model is not fitted yet!")
+        preds = [m.predict(dataset, segment=segment) for m in self._models]
+        if len(preds) == 1:
+            return preds[0]
+        frame = pd.concat(preds, axis=1)
+        return frame.mean(axis=1)

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

@@ -0,0 +1,200 @@
+"""LGBModel variant that early-stops on cross-sectional RankIC instead of l2.
+
+Standard qlib ``LGBModel`` early-stops on the regression loss (mse). For
+cross-sectional alpha signals the quantity we actually care about is the per-day
+rank correlation (Rank IC), which mse early-stopping does not optimize for.
+Experiments on the 50-ETF lake (SP-5d 55-feature panel) show that early-stopping
+on a custom RankIC feval lifts RankIC 0.047 -> 0.075 vs. the mse-stopped model.
+
+This class reuses ``LGBModel``'s data preparation but:
+
+  - tags each ``lgb.Dataset`` with per-day query ``group`` sizes so a ranking
+    metric can be computed per trading day;
+  - injects a custom ``feval`` (mean per-day Spearman of pred vs label) into
+    ``lgb.train``; early stopping then selects the iteration that maximizes
+    RankIC on the valid set;
+  - forces ``metric='None'`` + ``first_metric_only=True`` so early-stopping
+    tracks RankIC only (not the regression loss).
+
+Wired into a workflow yaml like:
+
+    model:
+        class: RankICLGBModel
+        module_path: tac_qlib.contrib.model.rank_gbdt
+        kwargs:
+            loss: mse
+            learning_rate: 0.03
+            num_leaves: 31
+            n_estimators: 500
+            ...
+
+The rank feval is used for early-stopping selection only; the objective stays
+the configured loss (default mse). Set ``rank_eval=False`` to fall back to the
+plain LGBModel behaviour (early-stop on the loss).
+
+Generic: works for any cross-sectional panel whose qlib dataset index has a
+``datetime`` level (each level value = one query group). The per-day groups are
+derived automatically, so no universe-specific configuration is needed.
+"""
+
+from __future__ import annotations
+
+from typing import List, Optional, Tuple
+
+import numpy as np
+import pandas as pd
+import lightgbm as lgb
+
+from qlib.data.dataset import DatasetH
+from qlib.data.dataset.handler import DataHandlerLP
+from qlib.contrib.model.gbdt import LGBModel
+from qlib.workflow import R
+
+__all__ = ["RankICLGBModel", "rankic_feval"]
+
+
+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.
+    """
+    if group is None or len(group) == 0:
+        return 0.0
+    offs = np.concatenate([[0], np.cumsum(group.astype(int))])
+    vals = []
+    for i in range(len(group)):
+        s = slice(offs[i], offs[i + 1])
+        p, l = preds[s], labels[s]
+        if len(p) < 3 or np.std(p) == 0 or np.std(l) == 0:
+            continue
+        vals.append(np.corrcoef(pd.Series(p).rank(), pd.Series(l).rank())[0, 1])
+    return float(np.mean(vals)) if vals else 0.0
+
+
+def rankic_feval(preds, dataset):
+    """LightGBM feval: mean RankIC (higher is better in lgb convention)."""
+    labels = dataset.get_label()
+    group = dataset.get_group()
+    ric = _per_day_spearman(preds, labels, group)
+    return "rankic", ric, True  # (name, value, higher_is_better)
+
+
+class RankICLGBModel(LGBModel):
+    """LGBModel that early-stops on per-day RankIC via a custom feval."""
+
+    def __init__(self, rank_eval: bool = True, **kwargs):
+        super().__init__(**kwargs)
+        self.rank_eval = rank_eval
+
+    def _prepare_data(self, dataset: DatasetH, reweighter=None) -> List[Tuple[lgb.Dataset, str]]:
+        ds_l = []
+        assert "train" in dataset.segments
+        for key in ["train", "valid"]:
+            if key in dataset.segments:
+                df = dataset.prepare(key, col_set=["feature", "label"], data_key=DataHandlerLP.DK_L)
+                if df.empty:
+                    raise ValueError("Empty data from dataset, please check your dataset config.")
+                x, y = df["feature"], df["label"]
+                if y.values.ndim == 2 and y.values.shape[1] == 1:
+                    y = np.squeeze(y.values)
+                else:
+                    raise ValueError("LightGBM doesn't support multi-label training")
+
+                if reweighter is None:
+                    w = None
+                elif hasattr(reweighter, "reweight"):
+                    w = reweighter.reweight(df)
+                else:
+                    raise ValueError("Unsupported reweighter type.")
+
+                # per-day query groups: each trading day is one group
+                if self.rank_eval and isinstance(df.index, pd.MultiIndex) and "datetime" in df.index.names:
+                    group = df.groupby(level="datetime").size().to_numpy(dtype=np.int32)
+                else:
+                    group = None
+
+                d = lgb.Dataset(x.values, label=y, weight=w, group=group, free_raw_data=False)
+                ds_l.append((d, key))
+        return ds_l
+
+    def _train_from_datasets(
+        self,
+        ds_l: List[Tuple[lgb.Dataset, str]],
+        num_boost_round: Optional[int] = None,
+        early_stopping_rounds: Optional[int] = None,
+        verbose_eval: int = 20,
+        evals_result=None,
+        **kwargs,
+    ) -> Tuple[lgb.Booster, dict, List[str]]:
+        """Train a Booster from already-prepared ``lgb.Dataset`` objects.
+
+        Pure training — no ``R.log_metrics`` — so it can be called from worker
+        threads (qlib's ``R`` recorder is not thread-safe; the caller decides
+        when/where to log). Returns ``(booster, evals_result, segment_names)``.
+        """
+        if evals_result is None:
+            evals_result = {}
+        ds, names = list(zip(*ds_l))
+
+        callbacks = [
+            lgb.early_stopping(
+                self.early_stopping_rounds if early_stopping_rounds is None else early_stopping_rounds
+            ),
+            lgb.log_evaluation(period=verbose_eval),
+            lgb.record_evaluation(evals_result),
+        ]
+        if self.rank_eval:
+            # early-stopping must be driven ONLY by the RankIC feval, not l2.
+            # metric='None' suppresses the default l2 metric; first_metric_only
+            # makes early_stopping track the single remaining (rankic) metric.
+            self.params["metric"] = "None"
+            self.params["first_metric_only"] = True
+            feval = rankic_feval
+        else:
+            self.params.pop("metric", None)
+            self.params.pop("first_metric_only", None)
+            feval = None
+
+        booster = lgb.train(
+            self.params,
+            ds[0],
+            num_boost_round=self.num_boost_round if num_boost_round is None else num_boost_round,
+            valid_sets=ds,
+            valid_names=names,
+            feval=feval,
+            callbacks=callbacks,
+            **kwargs,
+        )
+        return booster, evals_result, list(names)
+
+    def _log_evals(self, evals_result, names: List[str], prefix: str = "") -> None:
+        """Log recorded evaluation curves to qlib's active recorder."""
+        for k in names:
+            for key, val in evals_result.get(k, {}).items():
+                name = f"{prefix}{key}.{k}"
+                for epoch, m in enumerate(val):
+                    R.log_metrics(**{name.replace("@", "_"): m}, step=epoch)
+
+    def fit(
+        self,
+        dataset: DatasetH,
+        num_boost_round: Optional[int] = None,
+        early_stopping_rounds: Optional[int] = None,
+        verbose_eval: int = 20,
+        evals_result=None,
+        reweighter=None,
+        **kwargs,
+    ):
+        if evals_result is None:
+            evals_result = {}
+        ds_l = self._prepare_data(dataset, reweighter)
+        self.model, evals_result, names = self._train_from_datasets(
+            ds_l,
+            num_boost_round=num_boost_round,
+            early_stopping_rounds=early_stopping_rounds,
+            verbose_eval=verbose_eval,
+            evals_result=evals_result,
+            **kwargs,
+        )
+        self._log_evals(evals_result, names)

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

@@ -0,0 +1,3 @@
+from .optimal_stop import OptimalStopControl  # noqa: F401
+
+__all__ = ["OptimalStopControl"]

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

@@ -0,0 +1,217 @@
+"""Optimal-stopping / stochastic-control strategy for cross-sectional signals.
+
+Entry is a control policy: a symbol opens a position only when its cross-sectional
+signal percentile is at or above ``entry_pct`` (i.e. it is one of the top-ranked
+names) and the portfolio has fewer than ``topk`` open positions.
+
+Exit is an optimal-stopping rule: a held position is stopped (closed) when its
+signal percentile falls below ``exit_pct`` (the continuation value of holding is
+no longer worth the risk), OR after ``max_hold_days`` (time stop / finite
+horizon), OR when the position P&L breaches ``sl`` (loss control) and the
+position has been held at least ``min_hold_days``.
+
+Sizing is fixed ``notional`` per position (equal-weight control), unlike the
+TopkDropout cash-allocation heuristic.
+
+Wired into qrun workflows like any ``BaseStrategy`` (see ``PortAnaRecord``
+config). Mirrors the API usage of qlib's ``TopkDropoutStrategy``: ``Order``/
+``OrderDir`` from ``qlib.backtest.decision``, ``trade_calendar`` /
+``trade_exchange`` / ``trade_position`` injected by the backtest executor.
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+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__ = ["OptimalStopControl"]
+
+DEFAULT_NOTIONAL = 20_000.0
+DEFAULT_ENTRY_PCT = 0.80
+DEFAULT_EXIT_PCT = 0.50
+DEFAULT_MAX_HOLD_DAYS = 10
+DEFAULT_MIN_HOLD_DAYS = 2
+DEFAULT_SL = -0.06
+
+
+class OptimalStopControl(BaseSignalStrategy):
+    """Optimal-stopping long-only strategy over a cross-sectional signal.
+
+    Parameters
+    ----------
+    topk : max number of concurrent positions.
+    entry_pct : min cross-sectional score percentile required to OPEN (0..1).
+    exit_pct : held positions are stopped when score percentile < exit_pct.
+    max_hold_days : hard time stop (finite-horizon close).
+    min_hold_days : minimum holding days before stop-loss is evaluated.
+    notional : $ per position (equal-weight control).
+    sl : stop-loss threshold as fraction of entry price (<= 0), disabled if 0.
+    """
+
+    def __init__(
+        self,
+        *,
+        signal=None,
+        topk: int = 10,
+        entry_pct: float = DEFAULT_ENTRY_PCT,
+        exit_pct: float = DEFAULT_EXIT_PCT,
+        max_hold_days: int = DEFAULT_MAX_HOLD_DAYS,
+        min_hold_days: int = DEFAULT_MIN_HOLD_DAYS,
+        notional: float = DEFAULT_NOTIONAL,
+        sl: float = DEFAULT_SL,
+        risk_degree: float = 0.95,
+        trade_exchange=None,
+        level_infra=None,
+        common_infra=None,
+        **kwargs,
+    ):
+        super().__init__(
+            signal=signal,
+            trade_exchange=trade_exchange,
+            level_infra=level_infra,
+            common_infra=common_infra,
+            **kwargs,
+        )
+        self.topk = topk
+        self.entry_pct = entry_pct
+        self.exit_pct = exit_pct
+        self.max_hold_days = max_hold_days
+        self.min_hold_days = min_hold_days
+        self.notional = notional
+        self.sl = sl
+
+    # ------------------------------------------------------------------ utils
+    @staticmethod
+    def _pct_rank(score: pd.Series) -> pd.Series:
+        return score.rank(pct=True)
+
+    def _entry_price(self, pos) -> float:
+        # Position stores avg entry price under key "price" (see Position.position)
+        price = pos.position.get("price")
+        if price is None:
+            price = pos.get_stock_amount("price")
+        return float(price)
+
+    def _pnl_pct(self, pos, mark: float) -> float:
+        entry = self._entry_price(pos)
+        if not entry or entry != entry:
+            return 0.0
+        return mark / entry - 1.0
+
+    def _is_tradable(self, code, start, end, direction) -> bool:
+        try:
+            return self.trade_exchange.is_stock_tradable(
+                stock_id=code, start_time=start, end_time=end, direction=direction
+            )
+        except TypeError:  # some exchanges take no direction kwarg
+            return self.trade_exchange.is_stock_tradable(stock_id=code, start_time=start, end_time=end)
+
+    # ------------------------------------------------------------ decision
+    def generate_trade_decision(self, execute_result=None):
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start, trade_end = self.trade_calendar.get_step_time(trade_step)
+        pred_start, pred_end = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start, end_time=pred_end)
+        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)
+
+        pct = self._pct_rank(pred_score)
+        time_per_step = self.trade_calendar.get_freq()
+        current_temp = __import__("copy").deepcopy(self.trade_position)
+
+        holdings = {}
+        for code in current_temp.get_stock_list():
+            if abs(current_temp.get_stock_amount(code)) > 1e-6:
+                holdings[code] = current_temp
+
+        # ---- optimal stopping: close held positions -----------------------
+        sell_orders: List[Order] = []
+        closed_today = set()
+        kept = {}
+        for code, pos in holdings.items():
+            held = current_temp.get_stock_count(code, bar=time_per_step)
+            mark = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start, end_time=trade_end, direction=Order.SELL
+            )
+            if mark is None or mark != mark:
+                continue
+            rank = pct.get(code, 0.0)
+            stop_pnl = held >= self.min_hold_days and self.sl < 0 and self._pnl_pct(pos, mark) <= self.sl
+            if held >= self.max_hold_days or rank < self.exit_pct or stop_pnl:
+                amt = abs(current_temp.get_stock_amount(code))
+                o = Order(stock_id=code, amount=amt, start_time=trade_start,
+                          end_time=trade_end, direction=Order.SELL)
+                if self.trade_exchange.check_order(o):
+                    sell_orders.append(o)
+                    self.trade_exchange.deal_order(o, position=current_temp)
+                closed_today.add(code)
+            else:
+                kept[code] = mark
+
+        # ---- equal-weight control: target notional per name -----------------
+        # candidate opens: top-ranked names whose signal pct >= entry_pct
+        rank_desc = pred_score.sort_values(ascending=False)
+        held_codes = set(kept)
+        opens = []
+        for sym in rank_desc.index:
+            if len(opens) >= self.topk:
+                break
+            if sym in held_codes:
+                continue
+            if pct.get(sym, 0.0) < self.entry_pct:
+                continue
+            if not self._is_tradable(sym, trade_start, trade_end, OrderDir.BUY):
+                continue
+            opens.append(sym)
+
+        targets = held_codes | set(opens)
+        if not targets:
+            return TradeDecisionWO(sell_orders, self)
+
+        # total value (cash + marked positions) -> per-target notional
+        total_value = current_temp.get_cash()
+        for code, mark in kept.items():
+            total_value += abs(current_temp.get_stock_amount(code)) * mark
+
+        target_notional = total_value * self.risk_degree / max(1, len(targets))
+
+        # ---- rebalance kept positions toward target weight ------------------
+        buy_orders: List[Order] = []
+        for code, mark in kept.items():
+            cur = abs(current_temp.get_stock_amount(code)) * mark
+            diff_notional = target_notional - cur
+            if abs(diff_notional) / target_notional < 0.02:
+                continue  # skip tiny rebalances
+            amount_delta = diff_notional / mark
+            direction = Order.BUY if amount_delta > 0 else Order.SELL
+            o = Order(stock_id=code, amount=abs(amount_delta), start_time=trade_start,
+                      end_time=trade_end, direction=direction)
+            if self.trade_exchange.check_order(o):
+                (buy_orders if direction == Order.BUY else sell_orders).append(o)
+                self.trade_exchange.deal_order(o, position=current_temp)
+
+        # ---- open new positions at target weight ----------------------------
+        for sym in opens:
+            px = self.trade_exchange.get_deal_price(
+                stock_id=sym, start_time=trade_start, end_time=trade_end, direction=OrderDir.BUY
+            )
+            if px is None or px != px or px <= 0:
+                continue
+            amount = target_notional / px
+            factor = self.trade_exchange.get_factor(
+                stock_id=sym, start_time=trade_start, end_time=trade_end
+            )
+            amount = self.trade_exchange.round_amount_by_trade_unit(amount, factor)
+            o = Order(stock_id=sym, amount=amount, start_time=trade_start,
+                      end_time=trade_end, direction=Order.BUY)
+            if self.trade_exchange.check_order(o):
+                buy_orders.append(o)
+
+        return TradeDecisionWO(sell_orders + buy_orders, self)

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

@@ -0,0 +1,25 @@
+from .config import (
+    LakeConfig,
+    BAR_FIELD_MAP,
+    FREQ_TO_TIMEFRAME,
+    UNKNOWN_FIELD_NAMES,
+    timeframe_for_freq,
+    resolve_lake_root,
+)
+from .providers import (
+    LakeCalendarProvider,
+    LakeInstrumentProvider,
+    LakeFeatureProvider,
+)
+
+__all__ = [
+    "LakeConfig",
+    "BAR_FIELD_MAP",
+    "FREQ_TO_TIMEFRAME",
+    "UNKNOWN_FIELD_NAMES",
+    "timeframe_for_freq",
+    "resolve_lake_root",
+    "LakeCalendarProvider",
+    "LakeInstrumentProvider",
+    "LakeFeatureProvider",
+]

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

@@ -0,0 +1,175 @@
+"""TradeAC lake configuration helpers.
+
+The lake is a hive-partitioned parquet store (see ``tac-engine/skills/tradeac-lake``):
+
+    $TAC_LAKE_DIR/
+    ├── market=US/
+    │   └── timeframe=1d/
+    │       └── symbol=AAPL.parquet        # OHLCV bars: t, date, o, h, l, c, v, n, vw
+    ├── features/                          # ta-lib indicators, wide format
+    │   └── market=US/
+    │       └── timeframe=1d/
+    │           └── symbol=AAPL.parquet    # t, sma_5, sma_20, rsi_14, ...
+    ├── calendar.parquet                   # trading days per market
+    ├── coverage.parquet                   # per (market,timeframe,symbol) loaded windows
+    └── symbols.parquet                    # asset master
+"""
+
+from __future__ import annotations
+
+import os
+from pathlib import Path
+from typing import Dict, List, Optional
+
+import pandas as pd
+
+#: qlib freq string (Freq.__str__) -> lake timeframe partition name
+FREQ_TO_TIMEFRAME: Dict[str, str] = {
+    "day": "1d",
+    "1d": "1d",
+    "min": "1m",
+    "1min": "1m",
+    "5min": "5m",
+    "10min": "10m",
+    "15min": "15m",
+    "30min": "30m",
+    "hour": "1h",
+    "1hour": "1h",
+    "2hour": "2h",
+    "4hour": "4h",
+    "week": "1w",
+    "1week": "1w",
+    "month": "1M",
+    "1month": "1M",
+}
+
+#: bar-field map: qlib field name (without the leading ``$``) -> lake bar column
+BAR_FIELD_MAP: Dict[str, str] = {
+    "open": "o",
+    "high": "h",
+    "low": "l",
+    "close": "c",
+    "volume": "v",
+    "vwap": "vw",
+    "avg_amount": "vw",  # amount / volume
+}
+
+#: fields that qlib core/backtest queries but the lake does not store -> all-NaN
+UNKNOWN_FIELD_NAMES = ("factor", "change", "trade_unit", "suspend_flag")
+
+#: columns in the parquet files that are not features
+NON_FEATURE_COLUMNS = ("t", "date", "market", "timeframe", "symbol")
+
+
+def timeframe_for_freq(freq: str) -> str:
+    """Map a qlib frequency (e.g. ``day``, ``1min``) to a lake timeframe (e.g. ``1d``)."""
+    f = str(freq).lower()
+    if f not in FREQ_TO_TIMEFRAME:
+        raise ValueError(
+            f"unsupported qlib freq {freq!r}; supported freqs: {sorted(set(FREQ_TO_TIMEFRAME))}"
+        )
+    return FREQ_TO_TIMEFRAME[f]
+
+
+def resolve_lake_root(lake_root: Optional[str] = None) -> Path:
+    """Resolve the lake root: explicit arg > ``TAC_LAKE_DIR`` (no fallback).
+
+    ``TAC_LAKE_DIR`` is **mandatory** — there is deliberately no default
+    A missing/empty value raises so a
+    misconfigured environment never silently points at a wrong directory.
+    """
+    if lake_root is None:
+        lake_root = os.environ.get("TAC_LAKE_DIR")
+    if not lake_root:
+        raise RuntimeError(
+            "TAC_LAKE_DIR is not set. Point it at the TradeAC lake root, e.g. "
+            "export TAC_LAKE_DIR=/home/data/lake (docker) or set an absolute "
+            "path in your local .env."
+        )
+    return Path(str(lake_root)).expanduser().resolve()
+
+
+class LakeConfig:
+    """Path helpers + cached readers for a (lake_root, market) combination."""
+
+    def __init__(self, lake_root: Optional[str] = None, market: str = "US"):
+        self.lake_root: Path = resolve_lake_root(lake_root)
+        self.market: str = (market or "US").upper()
+
+    # ---- paths --------------------------------------------------------------
+    def bar_dir(self, timeframe: str) -> Path:
+        return self.lake_root / f"market={self.market}" / f"timeframe={timeframe}"
+
+    def bar_path(self, timeframe: str, symbol: str) -> Path:
+        return self.bar_dir(timeframe) / f"symbol={str(symbol).upper()}.parquet"
+
+    def features_dir(self, timeframe: str) -> Path:
+        return self.lake_root / "features" / f"market={self.market}" / f"timeframe={timeframe}"
+
+    def features_path(self, timeframe: str, symbol: str) -> Path:
+        return self.features_dir(timeframe) / f"symbol={str(symbol).upper()}.parquet"
+
+    def calendar_path(self) -> Path:
+        return self.lake_root / "calendar.parquet"
+
+    def symbols_path(self) -> Path:
+        return self.lake_root / "symbols.parquet"
+
+    def coverage_path(self) -> Path:
+        return self.lake_root / "coverage.parquet"
+
+    # ---- metadata readers ----------------------------------------------------
+    def load_symbols(self) -> List[str]:
+        """All symbols known to the lake (from ``symbols.parquet``)."""
+        p = self.symbols_path()
+        if not p.exists():
+            return []
+        df = pd.read_parquet(p)
+        if "symbol" not in df.columns:
+            return []
+        return sorted(df["symbol"].astype(str).str.upper().tolist())
+
+    def symbol_spans(self, symbol: str, timeframe: str) -> List[tuple]:
+        """Listing span(s) ``[(start_iso, end_iso)]`` for a symbol from coverage.parquet."""
+        p = self.coverage_path()
+        if p.exists():
+            try:
+                df = pd.read_parquet(p)
+            except Exception:  # pragma: no cover - defensive
+                df = pd.DataFrame()
+            if len(df):
+                df = df[
+                    (df.get("market") == self.market)
+                    & (df.get("timeframe") == timeframe)
+                    & (df.get("symbol") == str(symbol).upper())
+                ]
+                if len(df):
+                    row = df.iloc[0]
+                    first = pd.Timestamp(row["first_t"]).date()
+                    last = pd.Timestamp(row["last_t"]).date()
+                    return [(first.isoformat(), last.isoformat())]
+        # fallback: derive from the bar file itself
+        p = self.bar_path(timeframe, symbol)
+        if p.exists():
+            import pyarrow.parquet as pq
+
+            tbl = pq.read_table(p, columns=["t"])
+            first = pd.Timestamp(tbl.column("t")[0].as_py()).date()
+            last = pd.Timestamp(tbl.column("t")[-1].as_py()).date()
+            return [(first.isoformat(), last.isoformat())]
+        return [("1970-01-01", "2099-12-31")]
+
+    def load_calendar_dates(self) -> List[pd.Timestamp]:
+        """Trading days (midnight timestamps) for the market, from ``calendar.parquet``."""
+        p = self.calendar_path()
+        if p.exists():
+            df = pd.read_parquet(p)
+            if "date" in df.columns:
+                if "market" in df.columns:
+                    df = df[df["market"] == self.market]
+                dates = pd.to_datetime(df["date"]).dt.normalize().sort_values().unique()
+                return [pd.Timestamp(x) for x in dates]
+        return []
+
+    def __repr__(self) -> str:  # pragma: no cover
+        return f"LakeConfig(lake_root={self.lake_root}, market={self.market})"

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

@@ -0,0 +1,231 @@
+"""qlib data providers backed by the TradeAC parquet lake.
+
+These providers plug into the standard qlib mechanism: ``qlib.init(calendar_provider=...,
+instrument_provider=..., feature_provider=...)`` instantiates them and binds them to the
+``Cal`` / ``Inst`` / ``FeatureD`` wrappers (see ``qlib.data.data.register_all_wrappers``).
+The rest of qlib (``LocalDatasetProvider`` expression engine, backtest ``Exchange``) keeps
+working unchanged because the interface contract is identical to the file-based providers:
+
+- ``feature()`` returns a ``pd.Series`` indexed by the **calendar position** range
+  ``[start_index, end_index]`` (matching ``FileFeatureStorage.__getitem__`` semantics).
+- ``list_instruments()`` returns ``{symbol: [(start, end), ...]}``.
+- ``load_calendar()`` returns a list of ``pd.Timestamp`` trading days.
+"""
+
+from __future__ import annotations
+
+import bisect
+from typing import Dict, List, Optional, Union
+
+import numpy as np
+import pandas as pd
+
+from qlib.data.data import CalendarProvider, FeatureProvider, InstrumentProvider
+from qlib.log import get_module_logger
+
+from .config import (
+    BAR_FIELD_MAP,
+    LakeConfig,
+    UNKNOWN_FIELD_NAMES,
+    timeframe_for_freq,
+)
+
+logger = get_module_logger("tac_qlib.data.providers")
+
+
+def _day_freq(freq: str) -> bool:
+    return str(freq).lower() in ("day", "1d")
+
+
+def _calendar_keys(cal: List[pd.Timestamp], freq: str) -> pd.Index:
+    """Convert calendar timestamps into the same key space as the lake parquet."""
+    if _day_freq(freq):
+        return pd.Index([pd.Timestamp(x).date() for x in cal])
+    return pd.Index([pd.Timestamp(x) for x in cal])
+
+
+class LakeCalendarProvider(CalendarProvider):
+    """Trading calendar read from ``<lake>/calendar.parquet`` (fallback: derived from bars)."""
+
+    def __init__(self, lake_root: Optional[str] = None, market: str = "US"):
+        super().__init__()
+        self.cfg = LakeConfig(lake_root, market)
+
+    def load_calendar(self, freq, future):
+        timeframe = timeframe_for_freq(freq)
+        if not _day_freq(freq):
+            raise NotImplementedError(
+                f"freq={freq!r} (timeframe={timeframe}) is not supported yet: the lake calendar "
+                f"only covers daily sessions; add a minute-level calendar to `calendar.parquet`"
+            )
+
+        dates = self.cfg.load_calendar_dates()
+        if not dates:
+            # Fallback: derive the trading-day set from the persisted bar files.
+            bar_dir = self.cfg.bar_dir(timeframe)
+            if bar_dir.exists():
+                import pyarrow.parquet as pq
+
+                cal: Dict[pd.Timestamp, None] = {}
+                for p in sorted(bar_dir.glob("symbol=*.parquet")):
+                    tbl = pq.read_table(p, columns=["t"])
+                    for v in tbl.column("t"):
+                        cal[pd.Timestamp(v.as_py()).normalize()] = None
+                dates = sorted(cal.keys())
+            if not dates:
+                return []
+
+        if future:
+            # append the next calendar day so that "today" is a valid trade date
+            last = dates[-1]
+            dates = dates + [pd.Timestamp(last) + pd.Timedelta(days=1)]
+        return dates
+
+
+class LakeInstrumentProvider(InstrumentProvider):
+    """Instruments from ``<lake>/symbols.parquet`` with listing spans from ``coverage.parquet``."""
+
+    def __init__(
+        self,
+        lake_root: Optional[str] = None,
+        market: str = "US",
+        markets: Optional[Dict[str, list]] = None,
+    ):
+        super().__init__()
+        self.cfg = LakeConfig(lake_root, market)
+        #: optional named pools, e.g. ``{"sp500": ["AAPL", "MSFT"], "etf": ["SPY"]}``.
+        #: ``all`` / any unregistered name resolves to every symbol in the lake.
+        self.markets: Dict[str, list] = markets or {}
+
+    def _resolve_symbols(self, market: Union[str, list]) -> List[str]:
+        if isinstance(market, (list, tuple, pd.Index, np.ndarray)):
+            return [str(s).upper() for s in market]
+        if isinstance(market, str) and "," in market:
+            return [s.strip().upper() for s in market.split(",") if s.strip()]
+        if market in self.markets:
+            return [str(s).upper() for s in self.markets[market]]
+        return self.cfg.load_symbols()
+
+    def list_instruments(self, instruments, start_time=None, end_time=None, freq="day", as_list=False):
+        market = instruments["market"]
+        timeframe = timeframe_for_freq(freq)
+
+        symbols = self._resolve_symbols(market)
+        if not symbols:
+            if as_list:
+                return []
+            return {}
+
+        # clip listing spans to the queried window (mirror of LocalInstrumentProvider)
+        from qlib.data.data import Cal  # pylint: disable=C0415
+
+        cal = Cal.calendar(freq=freq)
+        start_time = pd.Timestamp(start_time or cal[0])
+        end_time = pd.Timestamp(end_time or cal[-1])
+
+        out: Dict[str, list] = {}
+        for symbol in symbols:
+            spans = []
+            for begin, end in self.cfg.symbol_spans(symbol, timeframe):
+                lo = max(start_time, pd.Timestamp(begin))
+                hi = min(end_time, pd.Timestamp(end))
+                if lo <= hi:
+                    spans.append((lo, hi))
+            if spans:
+                out[symbol] = spans
+
+        filter_pipe = instruments.get("filter_pipe") or []
+        for filter_config in filter_pipe:
+            from qlib.data import filter as F  # pylint: disable=C0415
+
+            filter_t = getattr(F, filter_config["filter_type"]).from_config(filter_config)
+            out = filter_t(out, start_time, end_time, freq)
+
+        if as_list:
+            return list(out)
+        return out
+
+
+class LakeFeatureProvider(FeatureProvider):
+    """Feature data from the lake parquet (OHLCV bars + pre-computed ta-lib features).
+
+    Field routing:
+        - ``$open/$high/$low/$close/$volume/$vwap``      -> bar parquet columns
+        - ``$amount`` (= v*vw), ``$avg_amount`` (= vw)  -> derived from bar parquet
+        - ``$factor/$change/...``                        -> all-NaN (not stored)
+        - anything else                                  -> a ta-lib column in the features parquet
+    """
+
+    def __init__(self, lake_root: Optional[str] = None, market: str = "US"):
+        super().__init__()
+        self.cfg = LakeConfig(lake_root, market)
+        self._bar_cache: Dict[tuple, pd.DataFrame] = {}
+        self._feature_cache: Dict[tuple, pd.DataFrame] = {}
+
+    # ------------------------------------------------------------------ caches
+    def _load_bar_df(self, instrument: str, timeframe: str) -> pd.DataFrame:
+        key = (instrument, timeframe)
+        if key not in self._bar_cache:
+            p = self.cfg.bar_path(timeframe, instrument)
+            self._bar_cache[key] = pd.read_parquet(p) if p.exists() else pd.DataFrame()
+        return self._bar_cache[key]
+
+    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] = pd.read_parquet(p) if p.exists() else pd.DataFrame()
+        return self._feature_cache[key]
+
+    @staticmethod
+    def _keys(df: pd.DataFrame, freq: str) -> pd.Index:
+        ts = pd.to_datetime(df["t"])
+        return ts.dt.date if _day_freq(freq) else ts
+
+    # ------------------------------------------------------------------ fields
+    def _extract(self, instrument: str, field: str, timeframe: str, freq: str) -> Optional[pd.Series]:
+        """Return the field as a Series keyed by date/timestamp (None if not present in the lake)."""
+        bar = self._load_bar_df(instrument, timeframe)
+
+        if field in BAR_FIELD_MAP:
+            col = BAR_FIELD_MAP[field]
+            if col in bar.columns:
+                return bar[col].astype(float).set_axis(self._keys(bar, freq))
+            return None
+        if field == "amount":
+            if "v" in bar.columns and "vw" in bar.columns:
+                return (bar["v"] * bar["vw"]).astype(float).set_axis(self._keys(bar, freq))
+            return None
+        if field in UNKNOWN_FIELD_NAMES:
+            return None
+
+        feat = self._load_feature_df(instrument, timeframe)
+        if field in feat.columns:
+            return feat[field].astype(float).set_axis(self._keys(feat, freq))
+        return None
+
+    # ------------------------------------------------------------------ api
+    def _get_calendar(self, freq: str) -> List[pd.Timestamp]:
+        from qlib.data.data import Cal  # pylint: disable=C0415
+
+        cal = Cal.calendar(freq=freq)
+        return list(cal)
+
+    def feature(self, instrument, field, start_index, end_index, freq):
+        field = str(field)[1:]
+        timeframe = timeframe_for_freq(freq)
+
+        cal = self._get_calendar(freq)
+        n = len(cal)
+        lo = max(0, int(start_index))
+        hi = min(n - 1, int(end_index))
+        if lo > hi:
+            return pd.Series(dtype=np.float32)
+
+        keys = _calendar_keys(cal[lo : hi + 1], freq)
+        ser = self._extract(str(instrument).upper(), field, timeframe, freq)
+        if ser is None:
+            vals = np.full(len(keys), np.nan, dtype=np.float64)
+        else:
+            vals = ser.reindex(keys).to_numpy(dtype=np.float64)
+        return pd.Series(vals, index=pd.RangeIndex(lo, hi + 1))

workflows/README-moments.md

@@ -1,20 +0,0 @@
-# exp/10 sp5d-moment-features
-
-Variant C: generic-only 19 + 16 new moment/volatility families (skew, kurt,
-DSV+ratios, max_up/down, rv_ac1, rv_cv_22, sig lag-5). 35 sp_* fields, ou/hmm excluded.
-
-Run a3f7d1d40c3d4b839314fcf5b40f9b08 (tac-rd-moments / exp 12) — FINISHED.
-
-## Result: NEGATIVE (regression vs generic-only baseline)
-
-| Metric | generic-only 19 (run 7b1e797) | +moments 35 (run a3f7d1d) |
-|---|---|---|
-| Rank IC   | 0.0635 | 0.0466 |
-| Rank ICIR | 0.276  | 0.183  |
-| L-S Sharpe | 2.55  | 1.44  |
-| net excess (cost) | +3.1% IR 0.28 | -16.2% IR -1.57 |
-| MDD | -7.3% | -11.1% |
-
-Same failure mode as ou/hmm in exp 9: adding cross-sectional moment features
-to the 50-name panel degrades the rank signal. Generic-only 19 remains the
-best configuration. No further moment-family variants planned.

workflows/exp12_isolation_ensemble.yaml

@@ -1,22 +1,24 @@
 # -----------------------------------------------------------------------------
-# VARIANT C (generic + moments): keeps the winning generic-only 19-field set
+# ISOLATION: multi-seed RankIC ensemble, ablate-B generic-only feature set.
-# (jump,har,trend,hurst,signature) and adds the NEW generic moment families the
+#
-# engine now exposes:
+# Isolates the ensemble effect on the SP-5d rank signal. Same panel, segments,
-#   - realized skewness / kurtosis      (sp_rskew_5, sp_rskew_22, sp_rkurt_5, sp_rkurt_22)
+# history (full backfilled 2016+) and feature set as the exp-9 ablate-B winner
-#   - downside semi-variance + ratios   (sp_dsv_1/5/22, sp_dsv_ratio_1/5/22)
+# (generic-only sp_* families: jump,har,trend,hurst,signature), but replaces the
-#   - signed max moves                  (sp_max_up, sp_max_down)
+# single RankICLGBModel with a 5-seed RankICEnsembleLGBModel (42,7,2026,99,123)
-#   - RV autocorr / vol-of-vol          (sp_rv_ac1, sp_rv_cv_22)
+# that averages per-day predictions.
-#   - longer-lag signature terms        (sp_sig_level2_*_5)
+#
-# Drops the model-specific ou/hmm families (they scored high in importance but
+# Differs from exp-15 (tac-rd-rank-ensemble, mlflow exp 15) ONLY by dropping the
-# hurt the rank dimension in the all-24 run). Same panel/model as baseline.
+# 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/ablate_generic_moments.yaml \
+#   rd_run_workflow config_path=experiments/workflows/exp12_isolation_ensemble.yaml \
-#       experiment_name=tac-rd-moments
+#       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_max_up,sp_max_down,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_rv_ac1,sp_rv_cv_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,sp_sig_level2_lead_lag_5,sp_sig_level2_lag_lead_5,sp_rskew_5,sp_rskew_22,sp_rkurt_5,sp_rkurt_22,sp_dsv_1,sp_dsv_5,sp_dsv_22,sp_dsv_ratio_1,sp_dsv_ratio_5,sp_dsv_ratio_22" %}
+{%- 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 }}"
@@ -45,13 +47,13 @@ qlib_init:
         class: MLflowExpManager
         module_path: qlib.workflow.expm
         kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
+            uri: "sqlite:///mlruns.db"
-            default_exp_name: "tac-rd-moments"
+            default_exp_name: "tac-rd-rank-ensemble-isolated"
 
 task:
     model:
-        class: RankICLGBModel
+        class: RankICEnsembleLGBModel
-        module_path: tac_qlib.contrib.model.rank_gbdt
+        module_path: tac_qlib.contrib.model.rank_ensemble
         kwargs:
             loss: mse
             learning_rate: 0.02
@@ -66,7 +68,7 @@ task:
             subsample_freq: 1
             reg_alpha: 0.1
             reg_lambda: 1.0
-            seed: 42
+            seeds: "42,7,2026,99,123"
 
     dataset:
         class: DatasetH
@@ -78,8 +80,8 @@ task:
                 kwargs:
                     instruments: "{{ UNIVERSE }}"
                     start_time: 2015-01-03
-                    end_time: 2026-08-10
+                    end_time: 2026-08-14
-                    fit_start_time: 2015-01-03
+                    fit_start_time: 2016-01-04
                     fit_end_time: 2025-09-01
                     freq: day
                     lake_root: "{{ LAKE }}"
@@ -98,7 +100,7 @@ task:
                         - class: Fillna
                           kwargs: {}
             segments:
-                train: [2015-01-03, 2025-09-01]
+                train: [2016-01-04, 2025-09-01]
                 valid: [2025-09-03, 2026-01-03]
                 test: [2026-01-04, 2026-08-10]