Commit Graph

  • 6fe2158511 feat: enhance precision optimization in model training main 21in7 2026-03-03 00:57:19 +09:00
  • 3613e3bf18 feat: update active LGBM parameters and training log with new metrics 21in7 2026-03-03 00:21:43 +09:00
  • fce4d536ea feat: implement HOLD negative sampling and stratified undersampling in ML pipeline 21in7 2026-03-03 00:13:42 +09:00
  • 74966590b5 feat: apply stratified undersampling to hyperparameter tuning 21in7 2026-03-03 00:09:43 +09:00
  • 6cd54b46d9 feat: apply stratified undersampling to training pipeline 21in7 2026-03-03 00:03:09 +09:00
  • 0af138d8ee feat: add stratified_undersample helper function 21in7 2026-03-02 23:58:15 +09:00
  • b7ad358a0a fix: make HOLD negative sampling tests non-vacuous 21in7 2026-03-02 23:45:10 +09:00
  • 8e56301d52 feat: add HOLD negative sampling to dataset_builder 21in7 2026-03-02 23:34:45 +09:00
  • 99fa508db7 feat: add CLAUDE.md and settings.json for project documentation and plugin configuration 21in7 2026-03-02 20:01:18 +09:00
  • eeb5e9d877 feat: add ADX filter to block sideways market entries 21in7 2026-03-02 19:55:12 +09:00
  • c8a2c36bfb feat: add ADX calculation to indicators 21in7 2026-03-02 19:47:18 +09:00
  • b8b99da207 feat: update training log and enhance ML filter functionality 21in7 2026-03-02 18:24:38 +09:00
  • 77590accf2 feat: add architecture documentation for CoinTrader 21in7 2026-03-02 18:02:05 +09:00
  • a8cba2cb4c docs: enhance README with detailed listenKey auto-renewal process and error handling 21in7 2026-03-02 16:43:45 +09:00
  • 52affb5532 feat: implement User Data Stream for real-time TP/SL detection and PnL tracking 21in7 2026-03-02 16:33:08 +09:00
  • 05ae88dc61 fix: remove manual listenKey mgmt, add symbol filter, fix reenter race condition 21in7 2026-03-02 16:31:40 +09:00
  • 6237efe4d3 docs: update README with User Data Stream TP/SL detection feature 21in7 2026-03-02 16:27:50 +09:00
  • 4e8e61b5cf fix: guard against None current_trade_side in _calc_estimated_pnl 21in7 2026-03-02 16:27:17 +09:00
  • 4ffee0ae8b feat: run UserDataStream in parallel with candle stream 21in7 2026-03-02 16:25:13 +09:00
  • 7e7f0f4f22 fix: restore entry_price and entry_quantity on position recovery 21in7 2026-03-02 16:24:27 +09:00
  • c4f806fc35 feat: add entry state tracking and _on_position_closed callback 21in7 2026-03-02 16:21:59 +09:00
  • 22f1debb3d fix: re-raise CancelledError in UserDataStream for proper task cancellation 21in7 2026-03-02 16:20:37 +09:00
  • 4f3183df47 feat: add UserDataStream with keepalive and reconnect loop 21in7 2026-03-02 16:17:38 +09:00
  • 223608bec0 refactor: remove duplicate pnl/notify from _close_position (handled by callback) 21in7 2026-03-02 16:16:25 +09:00
  • e72126516b feat: extend notify_close with close_reason, net_pnl, diff fields 21in7 2026-03-02 16:14:26 +09:00
  • 63c2eb8927 feat: add listenKey create/keepalive/delete methods to exchange 21in7 2026-03-02 16:11:33 +09:00
  • dcdaf9f90a chore: update active LGBM parameters and add new training log entry 21in7 2026-03-02 15:03:35 +09:00
  • 6d82febab7 feat: implement Active Config pattern for automatic param promotion 21in7 2026-03-02 14:56:42 +09:00
  • d5f8ed4789 feat: update default LightGBM params to Optuna best (trial #46, AUC=0.6002) 21in7 2026-03-02 14:52:41 +09:00
  • ce02f1335c feat: add run_optuna.sh wrapper script for Optuna tuning 21in7 2026-03-02 14:50:50 +09:00
  • 4afc7506d7 feat: connect Optuna tuning results to train_model.py via --tuned-params 21in7 2026-03-02 14:45:15 +09:00
  • caaa81f5f9 fix: add shebang and executable permission to tune_hyperparams.py 21in7 2026-03-02 14:41:13 +09:00
  • 8dd1389b16 feat: add Optuna Walk-Forward AUC hyperparameter tuning pipeline 21in7 2026-03-02 14:39:07 +09:00
  • 4c09d63505 feat: implement upsert functionality in fetch_history.py to accumulate OI/funding data 21in7 2026-03-02 14:16:09 +09:00
  • 0fca14a1c2 feat: auto-detect first run in train_and_deploy.sh (365d full vs 35d upsert) 21in7 2026-03-02 14:15:00 +09:00
  • 2f5227222b docs: add initial data setup instructions and OI accumulation strategy 21in7 2026-03-02 14:13:45 +09:00
  • 10b1ecd273 feat: fetch 35 days for daily upsert instead of overwriting 365 days 21in7 2026-03-02 14:13:16 +09:00
  • 016b13a8f1 fix: fill NaN in oi_change/funding_rate after concat when columns missing in existing parquet 21in7 2026-03-02 14:13:00 +09:00
  • 3c3c7fd56b feat: add upsert_parquet to accumulate OI/funding data incrementally 21in7 2026-03-02 14:09:36 +09:00
  • aa52047f14 fix: prevent OI API failure from corrupting _prev_oi state and ML features 21in7 2026-03-02 14:01:50 +09:00
  • b57b00051a fix: update test to force LONG signal so build_features is called 21in7 2026-03-02 13:57:08 +09:00
  • 3f4e7910fd docs: update README to reflect realtime OI/funding rate ML feature integration 21in7 2026-03-02 13:55:45 +09:00
  • dfd4990ae5 feat: fetch realtime OI and funding rate on candle close for ML features 21in7 2026-03-02 13:55:29 +09:00
  • 4669d08cb4 feat: build_features accepts oi_change and funding_rate params 21in7 2026-03-02 13:50:39 +09:00
  • 2b315ad6d7 feat: add get_open_interest and get_funding_rate to BinanceFuturesClient 21in7 2026-03-02 13:46:25 +09:00
  • 7a1abc7b72 chore: update python-binance dependency and improve error handling in BinanceFuturesClient 21in7 2026-03-02 13:24:27 +09:00
  • de2a402bc1 feat: enhance cancel_all_orders method to include Algo order cancellation 21in7 2026-03-02 02:15:49 +09:00
  • 684c8a32b9 feat: add Algo Order API support and update ML feature handling 21in7 2026-03-02 02:03:50 +09:00
  • c89374410e feat: enhance trading bot functionality and documentation 21in7 2026-03-02 01:51:53 +09:00
  • 9ec78d76bd feat: implement immediate re-entry after closing position on reverse signal 21in7 2026-03-02 01:34:36 +09:00
  • 725a4349ee chore: Update MLXFilter model deployment and logging with new training results and ONNX file management 21in7 2026-03-02 01:08:12 +09:00
  • 5e6cdcc358 fix: _on_candle_closed async 콜백 구조 수정 — asyncio.create_task 제거 21in7 2026-03-02 01:00:59 +09:00
  • 361b0f4e00 fix: Update TradingBot signal processing to handle NaN values and improve MLFilter ONNX session configuration 21in7 2026-03-02 00:47:17 +09:00
  • 031adac977 chore: .gitignore에 .DS_Store 추가 및 MLXFilter 훈련 로그 업데이트 21in7 2026-03-02 00:41:34 +09:00
  • 747ab45bb0 fix: test_reload_model 단언을 실제 동작(파일 없으면 폴백 상태)에 맞게 수정 21in7 2026-03-02 00:38:08 +09:00
  • 6fa6e854ca fix: test_reload_model _model → _lgbm_model 주입 방식으로 수정 21in7 2026-03-02 00:36:47 +09:00
  • 518f1846b8 fix: 기존 테스트를 현재 코드 구조에 맞게 수정 — MLFilter API, FEATURE_COLS 수, 버퍼 최솟값 반영 21in7 2026-03-02 00:36:13 +09:00
  • 3bfd1ca5a3 fix: test_mlx_filter _make_X를 FEATURE_COLS 기반으로 수정 — 피처 확장 후 input_dim 불일치 해소 21in7 2026-03-02 00:34:21 +09:00
  • 7fdd8bff94 fix: MLXFilter self._mean/std 저장 전 nan_to_num 적용 — 전체-NaN 컬럼 predict_proba 오염 차단 21in7 2026-03-02 00:31:08 +09:00
  • bcc717776d fix: RS 계산을 np.divide(where=) 방식으로 교체 — epsilon 이상치 폭발 차단 21in7 2026-03-02 00:30:36 +09:00
  • 9cac8a4afd Merge pull request 'feat: OI nan 마스킹 / epsilon 통일 / 정밀도 우선 임계값' (#1) from feature/oi-nan-epsilon-precision-threshold into main gihyeon 2026-03-01 23:57:32 +09:00
  • 0f6a22fcb5 feat: MLX 임계값 탐색을 정밀도 우선(recall>=0.15 조건부)으로 변경 feature/oi-nan-epsilon-precision-threshold 21in7 2026-03-01 23:54:38 +09:00
  • aa413f4d7c feat: LightGBM 임계값 탐색을 정밀도 우선(recall>=0.15 조건부)으로 변경 21in7 2026-03-01 23:54:13 +09:00
  • 6ae0f9d81b fix: MLXFilter fit/predict에 nan-safe 정규화 적용 (nanmean + nan_to_num) 21in7 2026-03-01 23:53:49 +09:00
  • 820d8e0213 refactor: 분모 연산을 1e-8 epsilon 패턴으로 통일 21in7 2026-03-01 23:52:59 +09:00
  • 417b8e3c6a feat: OI/펀딩비 결측 구간을 np.nan으로 마스킹 (0.0 → nan) 21in7 2026-03-01 23:52:19 +09:00
  • 3b7ee3e890 chore: .worktrees/ gitignore에 추가 21in7 2026-03-01 23:50:18 +09:00
  • 24d3ba9411 feat: enhance data fetching and model training with OI and funding rate integration 21in7 2026-03-01 22:25:38 +09:00
  • 4245d7cdbf feat: implement 15-minute timeframe upgrade for model training and data processing 21in7 2026-03-01 22:16:15 +09:00
  • a6697e7cca feat: implement LightGBM model improvement plan with feature normalization and walk-forward validation 21in7 2026-03-01 22:02:32 +09:00
  • c6428af64e feat: enhance Jenkins pipeline with Discord notifications and model hot-reload functionality 21in7 2026-03-01 21:46:36 +09:00
  • d9238afaf9 feat: enhance MLX model training with combined data handling 21in7 2026-03-01 21:43:27 +09:00
  • db144750a3 feat: enhance model training and deployment scripts with time-weighted sampling 21in7 2026-03-01 21:25:06 +09:00
  • 301457ce57 chore: remove unused risk_per_trade references 21in7 2026-03-01 20:39:26 +09:00
  • ab580b18af feat: apply dynamic margin ratio in bot position sizing 21in7 2026-03-01 20:39:07 +09:00
  • 795689ac49 feat: replace risk_per_trade with margin_ratio in calculate_quantity 21in7 2026-03-01 20:38:18 +09:00
  • fe9690698a feat: add get_dynamic_margin_ratio to RiskManager 21in7 2026-03-01 20:37:46 +09:00
  • 95abac53a8 feat: add dynamic margin ratio config params 21in7 2026-03-01 20:37:04 +09:00
  • ac84fafbd0 feat: update Jenkinsfile for improved CI/CD process 21in7 2026-03-01 19:51:17 +09:00
  • 94d8cb352e docs: update README to include a motivational message for users 21in7 2026-03-01 19:37:16 +09:00
  • d1af736bfc feat: implement BTC/ETH correlation features for improved model accuracy 21in7 2026-03-01 19:30:17 +09:00
  • c4062c39d3 feat: add duplicate training log entry for model evaluation 21in7 2026-03-01 18:55:26 +09:00
  • 6e73df196c chore: exclude mlx from container requirements (Apple Silicon only) 21in7 2026-03-01 18:54:38 +09:00
  • de933b97cc feat: remove in-container retraining, training is now mac-only 21in7 2026-03-01 18:54:00 +09:00
  • fd96055e73 perf: replace generate_dataset with vectorized version in train_mlx_model 21in7 2026-03-01 18:53:21 +09:00
  • db134c032a perf: replace generate_dataset with vectorized version in train_model 21in7 2026-03-01 18:52:56 +09:00
  • e1560f882b feat: add vectorized dataset builder (1x pandas_ta call) 21in7 2026-03-01 18:52:34 +09:00
  • 8f834a1890 feat: implement training and deployment pipeline for LightGBM model on Mac to LXC 21in7 2026-03-01 18:30:01 +09:00
  • 298d4ad95e feat: enhance train_model.py to dynamically determine CPU count for parallel processing 21in7 2026-03-01 17:46:40 +09:00
  • b86c88a8d6 feat: add README and enhance scripts for data fetching and model training 21in7 2026-03-01 17:42:12 +09:00
  • 7e4e9315c2 feat: implement ML filter with LightGBM for trading signal validation 21in7 2026-03-01 17:07:18 +09:00
  • ce57479b93 ci: Jenkins pipeline + Gitea registry CI/CD 설정 21in7 2026-03-01 16:31:34 +09:00
  • 86904dddf7 fix: upgrade to Python 3.12 to support pandas-ta>=0.4.67b0 21in7 2026-03-01 16:14:03 +09:00
  • 60b32a978e chore: add Dockerfile, docker-compose.yml, .dockerignore 21in7 2026-03-01 16:07:20 +09:00
  • 3d05806155 feat: Discord 알림, 포지션 복구, 설정 개선 및 docs 추가 21in7 2026-03-01 15:57:08 +09:00
  • 117fd9e6bc fix: AsyncClient.futures_klines await 누락 수정 21in7 2026-03-01 13:04:02 +09:00
  • 4940de16fc feat: 봇 시작 시 과거 캔들 200개 프리로드 (즉시 신호 계산 가능) 21in7 2026-03-01 13:02:42 +09:00
  • 8e5ec82244 fix: BinanceSocketManager 메서드명 수정 (futures_kline_socket → kline_futures_socket) 21in7 2026-03-01 12:56:53 +09:00
  • a90618896d feat: 엔트리포인트 및 로깅 설정 완료, 테스트 mock 개선 21in7 2026-03-01 12:54:21 +09:00
  • 726e9cfd65 feat: TradingBot 메인 루프 구현 (진입/청산/손절/익절) 21in7 2026-03-01 12:53:40 +09:00