crd / faker_credit_score / 24617764513

100%

push

github

feat: add credit_score_profile() for correlated multi-bureau scores (#72)

* feat: add credit_score_profile() for correlated multi-bureau scores

Generate realistic credit scores for the same person across multiple
bureaus. A base score is chosen, then each bureau gets a small random
jitter (±25 points) to simulate real cross-bureau variance.

- Returns dict[str, CreditScoreResult] keyed by provider name
- Supports score_type, tier, and provider filtering
- Case-insensitive provider name matching
- Individual scores may drift across tier boundaries naturally
- Scores always clamped to model-valid ranges

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: apply code review findings

- Fix non-deterministic provider ordering by iterating the input list
  instead of a set, preserving caller's order for seeded reproducibility
- Tighten tier test assertion: 300 (not 275) is the true minimum since
  model range clamping prevents going below model_low
- Add edge case tests: empty providers, duplicate providers, provider
  filter combined with tier

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refactor: extract profile_jitter as class variable, derive test bounds

Replace hardcoded magic numbers in tests with values computed from
the source data (model ranges, tier boundaries, profile_jitter).
Tests now stay correct if jitter or tier boundaries change.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

16 of 16 branches covered (100.0%)

Branch coverage included in aggregate %.

84 of 84 relevant lines covered (100.0%)

1.0 hits per line