Testing and Validation

Testing is the quality gate for SPX models. Treat your model library like code: every change should be validated by a deterministic suite.

Recommended layers

1. MiL tests (pytest): drive SPX deterministically from Python and assert behavior end-to-end.

2. Scenario checks: cover multi-step sequences and fault transitions via Scenarios.

3. Snapshot fixtures: restore known-good starting states for stable regressions.

Minimal CI-shaped loop

docker compose up -d
pytest
docker compose down

See:

• MiL tests: Use in Unit Tests (MiL)

• CI: CI/CD Setup (GitHub Actions)

• Snapshots: Snapshots — Getting Started