4 Commits

Include renames

Author	SHA1	Message	Date
mathias	731673061e	fix(deps): histdata>=1.3 is unsatisfiable, pin to >=1.1 (newest on PyPI) ... The Autoresearch Loop workflow failed at "Set up Python venv" on every run: ERROR: Could not find a version that satisfies the requirement histdata>=1.3 (from versions: 1.0, 1.1) histdata never published 1.3 — 1.1 is the newest release. Lower the floor so pip can resolve and the loop can actually start.	2026-06-29 18:41:07 +00:00
mathiasandClaude Sonnet 4.6	68bf8f15c5	feat(eval): VaR breach rate metric (#12) + HMM regime detector (#13) — rq-04 prep ... #12 — VaR_breach_rate_99_oos_regime_cond metric: - internal/eval/var.go: VaRBreachRate() + kupiecPOF() + LinearProbePredict() (stdlib math only) - internal/eval/var_test.go: 8 golden tests (zero/all breach, perfect calibration, boundary) - cmd/eval/main.go: -metric var flag (no-leakage probe → VaR → Kupiec P) - scripts/var_breach.py: Python equivalent with METRIC_KEY constant (13 TDD tests) - train.py LOCKED VaR EVAL BLOCK: writes VaR_breach_rate_99_oos_regime_cond + kupiec_p to metrics.json - Fixed bug: train.py used bare 'os' before import; now uses module-level '_os' consistently #13 — HMM regime detector + JEPA conditioning seam: - scripts/prepare_regime.py: GaussianHMM (diag, 3-state) on realized_vol; states sorted by mean vol (0=calm, 1=stressed, 2=crisis); deterministic (random_state=42); outputs eurusd_regime.parquet - tests/test_regime.py: 11 TDD tests (dtype, states, determinism, vol sort, daily fallback) - train.py: JEPA_ENABLE_REGIME toggle + REGIME CONDITIONING SEAM (concat baseline, agent-editable) - requirements.txt: hmmlearn>=0.3, scikit-learn>=1.4 78 Python + all Go tests green. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-06-27 10:35:10 +02:00
mathiasandClaude Opus 4.8	485fdaa9f9	feat(data): EUR/USD M1 fetch + daily realized-vol prep (toy slice, #2/#11) ... fetch_data.py politely pulls EUR/USD M1 from histdata.com (maintained package handles the anti-hotlink token; per-year, spaced). prepare_data.py (LOCKED per Phase-1 contract) parses M1 -> daily series with realized_vol = the val_vol_r2 target (sqrt sum of squared intraday returns). Verified on 2019-2021: 937 days, March-2020 COVID RV spike 5.1x over 2019 median — real signal, target works. Data gitignored (DVC/MinIO = #10). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-23 16:49:59 +02:00
mathiasandClaude Opus 4.8	df910e4336	chore(phase0): reproducible compute gate — torch cu130 + GPU smoke test ... scripts/check_gpu.py verifies PyTorch cu130 sees the koala Blackwell GPU (sm_120) and computes — the Phase-0 prerequisite before any autoresearch experiment. Verified green: torch 2.12.1+cu130, RTX 5070, GPU matmul OK. Note: koala GPU is shared with the llama-swap LLM stack — run the autoresearch agent on iguana/berget so the card stays free for train.py. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-23 16:43:01 +02:00