prop/.gitignore
Graham McIntire d5da0cec2b
feat(ml): add prop.compare task for algorithm-vs-ML-vs-reality calibration
`mix prop.compare` runs both scorers against a recent contact sample
and measures both against achieved contact distance. Each run appends
to priv/calibration/history.jsonl, so trends in alg-ML divergence and
algorithm-distance correlation surface over weeks of running.

When drift exceeds threshold (alg-ML RMSE > 8 score points, or current
correlation > 0.10 below the history median), the task writes a
recommendation pointing at the right remediation:

    mix recalibrate_scorer    # algorithm drift → refit weights
    mix propagation_train     # ML drift → retrain on the new algorithm

That is the feedback loop: measure → recommend → recalibrate/retrain →
loop. Operational rather than automatic, so a bad data window can't
silently corrupt the model.

Pure analysis lives in Microwaveprop.Propagation.Calibration with its
own unit tests; the Mix task only handles data loading, file I/O, and
console formatting.
2026-04-28 14:38:02 -05:00

65 lines
1.3 KiB
Text

# The directory Mix will write compiled artifacts to.
/_build/
# Environment variables (secrets)
.envrc
# If you run "mix test --cover", coverage assets end up here.
/cover/
# The directory Mix downloads your dependencies sources to.
/deps/
# Where 3rd-party dependencies like PHX://priv/static/
# temporary files for e.g. tests
/tmp/
# Ignore package tarball (built via "mix hex.build").
microwaveprop-*.tar
# Ignore assets that are produced by build tools.
/priv/static/assets/
# Ignore digested assets cache.
/priv/static/cache_manifest.json
# In case you use Node.js/Tailwind standalone binary
/assets/node_modules/
# Ignore database dumps
*.sql
*.psql
*.dump
# Dotenv files
.env
.env.*
# Kubernetes secret manifests (contain plaintext credentials)
/k8s/secret.yaml
/k8s/*-secret.yaml
# Trained ML model weights — tracked in git for Docker builds
# GRIB2 test fixtures (large binary files, downloaded on-demand)
/test/fixtures/grib2/*.grib2
# Dev-only propagation score binaries written by PropagationGridWorker
/priv/dev_scores/
# Calibration reports written by `mix prop.compare` — local artifacts only
/priv/calibration/
# Dialyzer PLT files
/priv/plts/
# Git worktrees
/.worktrees/
# OS files
.DS_Store
Thumbs.db
# Python bytecode caches (scripts/ has helper Python tools)
__pycache__/
*.pyc