Loads trained DQN/PPO checkpoints, runs greedy inference on walk-forward
test windows, computes Sharpe ratio, max drawdown, win rate, profit factor,
and total return per fold. Outputs a JSON evaluation report with aggregate
metrics and sanity checks (beats-random, action diversity, fold consistency).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add ml/examples/train_baseline.rs that trains DQN and PPO models using
expanding walk-forward windows on real Databento OHLCV data.
Features:
- CLI args via clap (--model, --epochs, --batch-size, --data-dir, etc.)
- Recursive .dbn.zst file discovery and OHLCV bar loading
- 51-dim feature extraction via extract_ml_features()
- Walk-forward window generation with NormStats per fold
- DQN training loop with epsilon-greedy, experience replay, early stopping
- PPO training loop with GAE, trajectory collection, early stopping
- PnL-based reward (BUY/SELL/HOLD)
- Safetensors checkpoint saving per fold
- NormStats JSON export for evaluation reproducibility
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Implements expanding-window walk-forward cross-validation for time-series
ML models, preventing lookahead bias by always evaluating on unseen future
data. Includes z-score NormStats computed from training splits only.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds `ml/examples/download_baseline.rs` that downloads 730 days of
Databento OHLCV-1m data in quarterly chunks for 4 CME futures symbols.
Features:
- Reads universe config from TOML (symbols, date range, dataset)
- Splits date range into calendar-quarter chunks (~90 days each)
- Resume support: skips existing non-empty files
- Uses `get_range_to_file` for streaming writes to .dbn.zst
- Dry-run mode with cost estimate ($0.12/symbol/day)
- Confirmation prompt (skippable with --yes)
- Per-file progress with timing and byte counts
- Failure-tolerant: logs errors and continues
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Ubuntu 22.04's protobuf-compiler (v3.12) lacks proto3 optional field
support. Install protoc v25.1 from GitHub releases instead, which
matches the protoc version available in Debian bookworm-based images
used by the other 5 services.
Validated: docker build completes successfully, binary starts with
ml_training_service --help.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
All 6 service Dockerfiles were missing newly-added workspace crates
(web-gateway, ctrader-openapi, foxhunt-deploy, broker_gateway_service),
causing cargo workspace resolution failures during Docker builds.
Changes across all Dockerfiles:
- Add COPY directives for web-gateway, ctrader-openapi, foxhunt-deploy,
broker_gateway_service (new workspace members since Dockerfiles written)
- Add SQLX_OFFLINE=true env and .sqlx cache copy where missing
- Add perl and make system deps (needed for OpenSSL build from source)
- Remove COPY migrations (dir excluded by .dockerignore, not needed)
- Expand broker_gateway_service from 3-crate to full workspace copy
Validated: docker build --check passes all 6, cargo check -p passes all 6.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Essential workflow fixes:
- ci.yml: add SQLX_OFFLINE=true, replace fictitious cargo subcommands
(test-unit, ci-lint, audit-deps, etc.) with real cargo commands,
remove broken 9-way test matrix, remove instrument-coverage RUSTFLAGS
- test.yml: add SQLX_OFFLINE=true to build-check job, fix conflicting
clippy flags (-D and -W on clippy::all), update JWT secret to 32+ chars
- compilation-guard.yml: add SQLX_OFFLINE=true, remove references to
non-existent paths (services/trading-engine, crates/common/types),
remove dangerous auto-commit-to-main, remove MIRI on missing packages
Deleted duplicates (justification):
- comprehensive_testing.yml: duplicate of comprehensive-testing.yml
(same purpose, underscore vs hyphen naming)
- production-deploy.yml: duplicate of production-deployment.yml
(both named "Production Deployment Pipeline", this one has stale paths)
- coverage-fixed.yml: duplicate of coverage.yml
(uses deprecated actions-rs/toolchain@v1 and actions/cache@v3)
- performance.yml: duplicate of benchmark_regression.yml
(both "Performance Regression Detection", this one less mature)
- quality-baseline.json: not a workflow, stale fake data (999 warnings)
- quality-metrics.json: not a workflow, stale fake data
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Proves the critical trading pipeline path works end-to-end:
OHLCV -> features -> ensemble -> prediction. Uses all 10 real candle
inference adapters (DQN, PPO, TFT, Mamba2, Liquid-CfC, TGGN, TLOB,
KAN, xLSTM, Diffusion) with random weights, 60 synthetic OHLCV bars,
and asserts confidence/action/model-count invariants. No external
dependencies (no DB, no Docker).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Spawn a background tokio task that feeds synthetic OHLCV bars into the
ensemble coordinator's per-symbol feature extractors. The extractors
require ~51 bars of warmup before they can produce real 51-dim feature
vectors; without this feed they remain cold and return zeros.
The task uses a random-walk price generator with realistic base prices
for the four baseline futures symbols (ES, NQ, ZN, 6E) and configurable
interval (MARKET_FEED_INTERVAL_MS, default 1s) and symbol list
(MARKET_FEED_SYMBOLS). In production this will be replaced by a
Databento or exchange data feed.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Register TGGN, TLOB, KAN, xLSTM, and Diffusion inference adapters
alongside the existing DQN, PPO, TFT, Mamba2, and Liquid-CfC.
Rebalance weights to 0.10 each (equal weighting across 10 models).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Sequence-buffered adapter with 3-layer MLP projection
(flat_dim -> hidden -> hidden/2 -> 1). Ring buffer collects
seq_len feature vectors, returns neutral prediction until full,
then flattens and projects through the MLP with sigmoid output
mapping to direction [-1,1] and confidence [0,1].
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove panic!() calls from test_futures_baseline_micro_mapping,
use map()+Some() pattern consistent with rest of test suite
- Make DatasetSpec::from_universe() accept a name parameter instead
of hardcoding "futures-baseline"
- Add doc comment to UniverseConfigMeta explaining dead_code fields
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Detailed TDD implementation plan for wiring all 10 ML models into
the ensemble coordinator. Covers 5 new inference adapters (KAN, TGGN,
xLSTM, TLOB, Diffusion), main.rs registration with equal weights,
market data wiring, E2E integration test, Docker/CI validation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Wire existing components into end-to-end pipeline: 5 missing inference
adapters, 10-model registration, market data feeding, integration test,
Docker build validation, and CI pipeline triage.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove duplicate AssetClass enum from data_pipeline module and replace
with a re-export from asset_selection, which is now the canonical
location. This ensures data_pipeline::AssetClass and
asset_selection::AssetClass are the same type, enabling direct
comparison and preventing subtle type mismatch bugs.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
10-task plan: unify AssetClass, add trading_symbol field, futures_baseline()
preset, TOML config loading, DatasetSpec::from_universe() wiring, data
reorganization into cache structure, stale data cleanup, manifest generation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Define futures baseline universe (ES, NQ, ZN, 6E) with micro
contract mapping for limited capital. Reorganize scattered test
data into cache structure. Download spec for 730 days of OHLCV-1m.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Use tokio::task::block_in_place for safe sync→async bridging
- Replace count(*) with count(1) in division context (QuestDB parser bug)
- Add coalesce() for stddev NULL handling (single-row edge case)
- Consolidate integration tests into single lifecycle test with DROP+CREATE
- All 4 QuestDB tests pass against real QuestDB 8.2.3
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace stub metrics provider with real QuestDBMetricsProvider that
queries QuestDB via PostgreSQL wire protocol (port 8812) for rolling
model Sharpe, win rates, confidence buckets, and ensemble metrics.
- Add QuestDB service to docker-compose.yml (8.2.3, ports 9009/8812/9003)
- Create questdb_metrics.rs with 5 SQL queries and graceful fallback
- Wire QuestDBMetricsProvider into main.rs (replaces StubMetricsProvider)
- Fix unused Instant import in feedback_loop.rs (#[cfg(test)] scope)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
FeedbackLoop runs as a background tokio task, periodically running
weight and gate optimization cycles with kill switch monitoring and
retraining trigger detection. Wired into main.rs with a stub
MetricsProvider until QuestDB is deployed.
7 new tests covering kill switch, freeze/unfreeze, retraining triggers.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Gate optimizer adjusts conviction gate thresholds based on win-rate per
confidence bucket with cooldown and kill switch safety rails. Model
registry provides lifecycle management (Candidate → Staging → Production
→ Archived) with InMemoryModelRegistry for testing. P&L attribution
decomposes realized trade P&L into per-model contributions using signal
alignment.
24 new tests across 3 modules.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Introduces the asset_selection module with AssetClass enum (Equity/ETF/Future),
UniverseAsset struct, and AssetUniverse with filtering, lookup, and a us_starter()
preset of 7 highly liquid US assets. Includes scorer/selector placeholder stubs.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adjusts model weights based on rolling Sharpe ratios with:
- EMA smoothing (alpha=0.1)
- Bounds: [0.05, 0.40] per model
- Max step: 0.03 per cycle
- 24h cooldown, 7-day grace period for new models
- Kill switch freeze/unfreeze
8 tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Non-critical path: if QuestDB is unavailable, metrics buffer locally
(up to 10,000 entries) and flush when connection is restored.
Feature-gated under `questdb` feature. 6 tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds optional ConvictionGateEvaluator field to EnsembleCoordinator.
When configured, predict() evaluates all 7 gates before returning.
If any gate rejects, returns HOLD with rejection metadata.
Includes quorum ratio, trading session, and regime volatility helpers.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
8-task plan across 3 phases: data pipeline (types, cache, manager,
prepared dataset), asset selection (universe, scorer, selector),
and integration wiring. 7 new files, ~51 tests estimated.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Covers automated data downloading/caching for ML training and
a tiered asset selection funnel (universe → predictability → regime → signal).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Three pillars: 7-gate conviction system, autonomous feedback loop
with kill switch, Rust-native model registry. QuestDB analytics layer.
~55 tests, 7 new files planned across ml/ and trading_service/.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
7-gate conviction system, autonomous feedback loop with kill switch,
and Rust-native model registry — backed by PostgreSQL + QuestDB.
Resolve merge conflicts in hyperopt/adapters/mod.rs and enhanced_ml.rs.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
