| name | ngat-tests |
| description | Maintain and expand test coverage under tests/ with focus on indicator correctness, aggregation behavior, and reliable separation of unit vs integration test paths. Use when adding modules, fixing regressions, or improving confidence in data/strategy pipelines. |
Ngat Tests
Overview
Strengthen correctness checks for indicators and timeframe aggregation while making tests deterministic and CI-friendly.
Submodules
tests/test_indicator_generator.py: pandas-ta parity checks, Redis-backed fetch/aggregate smoke coverage.- indirect dependency on
core/timeframe_generator_worker.py fetch and aggregate helpers.
Pros
- Validate indicator outputs against pandas-ta directly.
- Cover single-value and dataframe-style indicator results.
Cons
- Blend unit and environment-dependent integration concerns in one file.
- Depend on live Redis data for one test path.
- Miss coverage for broker/auth, API contracts, and frontend transformations.
Caveats
- Mark Redis-dependent tests as integration and skip when infrastructure is absent.
- Avoid asserting against unstable vendor indicator column naming without normalization.
- Ensure timezone-aware fixtures for aggregation tests.
Forward Features
- Split tests into
unit/, integration/, and contract/ groups.
- Add API response contract tests for
/candles, /indicators, /candles/latest.
- Add worker-loop tests with bounded iterations and fake clocks.
