init

cbpred/__init__.py

@@ -0,0 +1,4 @@
+"""Unified 8H drawdown prediction pipeline."""
+
+__version__ = "0.1.0"
+

cbpred/__main__.py

@@ -0,0 +1,138 @@
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+from .pipeline import (
+    VOLATILITY_RANK_THRESHOLD,
+    parse_time_ms,
+    predict_from_csv,
+    prepare_data,
+    run_training,
+)
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="Train a unified 8H futures drawdown model")
+    sub = parser.add_subparsers(dest="command", required=True)
+
+    prepare = sub.add_parser("prepare", help="filter symbols, call banbot, export complete 8H CSVs")
+    prepare.add_argument("--out-dir", type=Path, default=Path("data"))
+    prepare.add_argument("--banbot-dir", type=Path, default=Path("/home/ctyun/ban/banbot"))
+    prepare.add_argument("--banbot-config", type=Path, default=Path("/home/ctyun/ban/banbot/biz/config.yml"))
+    prepare.add_argument("--dsn", default=None)
+    prepare.add_argument("--db-type", choices=["auto", "questdb", "timescale"], default="auto")
+    prepare.add_argument("--exchange", default="binance")
+    prepare.add_argument("--end", dest="end_ms", default=None, help="UTC date/ISO timestamp or milliseconds")
+    prepare.add_argument("--history-months", type=int, default=36)
+    prepare.add_argument("--holdout-months", type=int, default=10)
+    prepare.add_argument("--min-age-days", type=int, default=730)
+    prepare.add_argument("--chunk-size", type=int, default=100)
+    prepare.add_argument("--banbot-cmd", default="go run .")
+    prepare.add_argument("--banbot-export-cmd", default="go run ./cmd/cbpred-export")
+    prepare.add_argument("--skip-download", action="store_true")
+
+    train = sub.add_parser("train", help="build three feature stages, tune on validation symbols, test once")
+    train.add_argument("--data-dir", type=Path, default=Path("data"))
+    train.add_argument("--artifact-dir", type=Path, default=Path("artifacts_signed"))
+    train.add_argument("--end", dest="end_ms", default=None)
+    train.add_argument("--test-months", type=int, default=10)
+    train.add_argument("--seed", type=int, default=42)
+    train.add_argument("--n-estimators", type=int, default=250)
+    train.add_argument("--n-jobs", type=int, default=-1)
+    train.add_argument("--volatility-rank-threshold", type=float, default=VOLATILITY_RANK_THRESHOLD)
+
+    run = sub.add_parser("run", help="prepare data and train the model")
+    for action in prepare._actions[1:]:
+        if action.dest not in {"help", "command"}:
+            run._add_action(action)
+    run.add_argument("--artifact-dir", type=Path, default=Path("artifacts_signed"))
+    run.add_argument("--test-months", type=int, default=10)
+    run.add_argument("--seed", type=int, default=42)
+    run.add_argument("--n-estimators", type=int, default=250)
+    run.add_argument("--n-jobs", type=int, default=-1)
+    run.add_argument("--volatility-rank-threshold", type=float, default=VOLATILITY_RANK_THRESHOLD)
+
+    predict = sub.add_parser(
+        "predict", help="predict signed robust future-low q20/q80 from an exported 8H CSV"
+    )
+    predict.add_argument("--artifact", type=Path, default=Path("artifacts_signed/model.joblib"))
+    predict.add_argument("--csv", type=Path, required=True)
+    predict.add_argument("--at", dest="at_ms", default=None)
+    predict.add_argument("--symbol", default=None, help="symbol used for the fixed training asset profile")
+    return parser
+
+
+def main() -> None:
+    args = build_parser().parse_args()
+    if args.command == "prepare":
+        result = prepare_data(
+            out_dir=args.out_dir,
+            banbot_dir=args.banbot_dir,
+            banbot_config=args.banbot_config,
+            dsn=args.dsn,
+            db_type=args.db_type,
+            exchange=args.exchange,
+            end_ms=parse_time_ms(args.end_ms) if args.end_ms else None,
+            history_months=args.history_months,
+            holdout_months=args.holdout_months,
+            min_age_days=args.min_age_days,
+            chunk_size=args.chunk_size,
+            skip_download=args.skip_download,
+            banbot_cmd=args.banbot_cmd,
+            banbot_export_cmd=args.banbot_export_cmd,
+        )
+        print(json.dumps({"selected": len(result["symbols"]), "excluded": len(result["excluded"])}, indent=2))
+    elif args.command == "train":
+        result = run_training(
+            data_dir=args.data_dir,
+            artifact_dir=args.artifact_dir,
+            end_ms=parse_time_ms(args.end_ms) if args.end_ms else None,
+            test_months=args.test_months,
+            seed=args.seed,
+            n_estimators=args.n_estimators,
+            n_jobs=args.n_jobs,
+            volatility_rank_threshold=args.volatility_rank_threshold,
+        )
+        print(json.dumps(result, indent=2, ensure_ascii=False))
+    elif args.command == "run":
+        prepare_data(
+            out_dir=args.out_dir,
+            banbot_dir=args.banbot_dir,
+            banbot_config=args.banbot_config,
+            dsn=args.dsn,
+            db_type=args.db_type,
+            exchange=args.exchange,
+            end_ms=parse_time_ms(args.end_ms) if args.end_ms else None,
+            history_months=args.history_months,
+            holdout_months=args.holdout_months,
+            min_age_days=args.min_age_days,
+            chunk_size=args.chunk_size,
+            skip_download=args.skip_download,
+            banbot_cmd=args.banbot_cmd,
+            banbot_export_cmd=args.banbot_export_cmd,
+        )
+        result = run_training(
+            data_dir=args.out_dir,
+            artifact_dir=args.artifact_dir,
+            end_ms=parse_time_ms(args.end_ms) if args.end_ms else None,
+            test_months=args.test_months,
+            seed=args.seed,
+            n_estimators=args.n_estimators,
+            n_jobs=args.n_jobs,
+            volatility_rank_threshold=args.volatility_rank_threshold,
+        )
+        print(json.dumps(result, indent=2, ensure_ascii=False))
+    else:
+        result = predict_from_csv(
+            artifact_path=args.artifact,
+            csv_path=args.csv,
+            at_ms=parse_time_ms(args.at_ms) if args.at_ms else None,
+            symbol=args.symbol,
+        )
+        print(json.dumps(result, indent=2, ensure_ascii=False))
+
+
+if __name__ == "__main__":
+    main()