grabowski
6ae67dec63
CI / Format & lint (push) Successful in 22s
CI / Test suite (push) Successful in 59s
Docs / Validate documentation (push) Successful in 25s
Rust / fmt, clippy, test, build (push) Successful in 10m2s
Security / Dependency vulnerabilities (push) Successful in 2m11s
Security / Static analysis (push) Successful in 22s
Security / License report (push) Successful in 1m37s
Phase 2b/2c of docs/RUST_REWRITE.md. With FORECAST_IN_COLLECTOR=true (shared .env) ping-collector, after each cycle that brought newer RID or HII readings and at least hourly: loads the last 336 h (RID + HII mirror fill), fetches the rain forecast, runs every exported model (rust/forecast), stores forecast_history and forecast_current, and sends the ntfy notifications. The Python API then serves forecast_current as /forecast (computing in-process only when it is older than FORECAST_MAX_AGE_SECONDS) and neither forecasts nor notifies. ntfy (collector/src/notify.rs): a port of src/notify.py with the same topics, texts, hysteresis, capacity guard and notification_state keys, checked word for word on 751 messages generated by Python (testdata/gen_notify_cases.py). The change: for stations whose ThaiWater twin is the same sensor, a reading newer than RID's drives the level alerts (marked as ThaiWater telemetry), and the feed counts as stale only when neither source has a reading for 3 h. RID alone lagging is reported once after RID_STALE_ALERT_HOURS (18). This ends the nightly "gauge feed stale" / "recovered" pairs caused by RID's late reports. Commands: `forecast [--json] [--save] [--now] [--rain-file]`, `notify --dry-run`. scripts/forecast_parity.py compares Python and the collector on the live database (same moment, same rain); locally 48/48 rows identical. install_collector.sh: models export + that check on every install, --cutover-forecast (refuses on a failed check) and --rollback-forecast. Trees are packed into 24-byte nodes: 70 MB peak with all 15 models (was 140). Forecast values are written to NUMERIC as shortest round-trip text, so they round as psycopg2's do. DB test schemas get a counter (parallel tests collided on Windows' coarse clock).
58 lines
3.3 KiB
Markdown
58 lines
3.3 KiB
Markdown
# CLAUDE.md
|
|
|
|
Guidance for AI coding agents working in this repository.
|
|
|
|
## What this is
|
|
|
|
Flood monitoring and forecasting for the Ping River, Chiang Mai. Public dashboard and
|
|
API at https://water.buildfor.life/ (never publish the server's private/Tailscale IP).
|
|
Production: one systemd unit on a small VPS, `/opt/thailand-water-monitor`, user
|
|
`water-monitor`, interpreter `.venv/bin/python` (uv-managed), updated by `git pull`.
|
|
|
|
## Rules
|
|
|
|
- Python 3.11 only. `uv sync --python 3.11`; run everything as `uv run ...`.
|
|
- `make format` (black 88 / isort black profile, config in pyproject.toml) before
|
|
committing; CI fails on formatting. `make test` must stay green — tests are
|
|
synthetic-data only, never add one that needs the DB or network.
|
|
- Timestamps everywhere are Asia/Bangkok wall-clock with no offset. The dashboard
|
|
parses them with `parseTs()` and renders with `timeZone: TZ`; keep it that way.
|
|
- Model changes go through the rolling-origin harness (`scripts/evaluate_variants.py`)
|
|
and are judged on first-alert LEAD and false alarms, not MAE. Record results, positive
|
|
or negative, in `docs/FLOOD_FORECASTING.md` section 5. Do not change what is deployed
|
|
(`rise_rain` / hgb-v3) without a harness result that beats it on lead.
|
|
- `train_all()` must never silently produce a gauge-only (v2) model; the guard that
|
|
raises `RainUnavailableError` stays.
|
|
- No `git add -A`: zero-byte shell-accident files (`#`, `$(wc`, ...) have been committed
|
|
before. Stage files by name.
|
|
- Do not add Co-Authored-By trailers.
|
|
- The dashboard is a single file, `src/static/dashboard.html`, EN + TH via the `t()`
|
|
table: every user-visible string needs both languages.
|
|
- Rust (`rust/`, see rust/README.md): `cargo fmt --all`, `cargo clippy --all-targets
|
|
-- -D warnings`, `cargo test` (database tests need `TEST_DATABASE_URL`). The Rust
|
|
collector writes the same tables as the Python collectors: a change to a table,
|
|
upsert or parser on one side must be made on the other until the Python side is
|
|
retired (check with a live run of both into two databases).
|
|
- The forecast is served by Rust (`rust/forecast`) from models trained in Python.
|
|
A change to `src/ml/features.py`, `predict.py`, the HII fill or the dam frame
|
|
must be ported in the same change: regenerate
|
|
`rust/forecast/testdata/gen_forecast_fixtures.py` and keep
|
|
`cargo test -p ping-forecast` green. Changing thresholds, stages, leads or
|
|
notify constants: `uv run python -m src.ml.export --config` (a test fails
|
|
otherwise) and rebuild the collector. Notify texts:
|
|
`rust/collector/testdata/gen_notify_cases.py`.
|
|
|
|
## Where things are
|
|
|
|
- `src/web_api.py` FastAPI app; `src/water_scraper_v3.py` RID collector;
|
|
`src/hii_collector.py` ThaiWater/HII; `src/ml/` features/train/evaluate/predict,
|
|
`rain.py` (Open-Meteo), `dam.py`, `hii_rain.py`.
|
|
- `scripts/retrain.sh` + `water-monitor-retrain.timer`: monthly retrain with staged
|
|
promote. `scripts/dev_proxy.py`: serve the working-copy dashboard against the live API.
|
|
- `rust/collector` (`ping-collector`): Rust data collector; with it deployed the
|
|
Python app runs `COLLECTION_ENABLED=false`. With `FORECAST_IN_COLLECTOR=true` it
|
|
also forecasts (`rust/forecast`) and sends ntfy; the API serves its
|
|
`forecast_current`. Rewrite plan: `docs/RUST_REWRITE.md`.
|
|
- `docs/FLOOD_FORECASTING.md` is the authoritative model write-up; `docs/DATA_SOURCES.md`
|
|
the source catalog.
|