Return reference distribution data from LR endpoints

[laurensWe]

Extends the LR endpoints to return the full reference distribution data alongside the log-LR value. Previously both endpoints returned only lr; they now also include KM/KNM scores, LLR values, and confidence intervals.

• Adds LRResult and LRStriationResult frozen dataclasses in controller.py to carry all output from process_lr_impression / process_lr_striation (previously returned float)
• Adds _extract_intervals helper to safely unpack llr_intervals (handles None)
• Extends LRResponse with km_scores, knm_scores, km_llr, knm_llr, and per-group CI fields; flat serializer updated to include the new fields
• Adds LRStriationResponse(LRResponse) with km_scores_transformed / knm_scores_transformed for the log-odds–transformed striation scores; calculate_lr_striation now declares this as its return type
• Updates assert_lr_response_valid to assert all new list fields are present in integration test responses
• Adjusts TestLRResponse fixtures and the _CONCRETE_CLASSES exclusion list to account for the schema change

[github-actions]

Diff Coverage

Diff: origin/main..HEAD, staged and unstaged changes

• packages/scratch-core/src/conversion/likelihood_ratio.py (93.2%): Missing lines 24,26,43
• packages/scratch-core/src/conversion/plots/data_formats.py (100%)
• packages/scratch-core/src/conversion/plots/utils.py (83.3%): Missing lines 398,403-404
• packages/scratch-core/src/conversion/utils.py (100%)
• src/extractors/schemas.py (100%)
• src/processors/controller.py (98.2%): Missing lines 51
• src/processors/router.py (94.7%): Missing lines 68
• src/processors/schemas.py (100%)

Summary

• Total: 190 lines
• Missing: 8 lines
• Coverage: 95%

packages/scratch-core/src/conversion/likelihood_ratio.py

  20 
  21     @model_validator(mode="after")
  22     def _validate_matching_lengths(self) -> Self:
  23         if len(self.km_scores) != len(self.km_llr_data.llrs):
! 24             raise ValueError("km_scores and km_lrs must have the same length")
  25         if len(self.knm_scores) != len(self.knm_llr_data.llrs):
! 26             raise ValueError("knm_scores and knm_lrs must have the same length")
  27         return self
  28 
  29     @property
  30     def scores(self) -> np.ndarray:

---

  39         """Concatenated KM and KNM LLR intervals, shape (n, 2)."""
  40         km = self.km_llr_data.llr_intervals
  41         knm = self.knm_llr_data.llr_intervals
  42         if km is None or knm is None:
! 43             raise ValueError("Only models with llr_intervals can be used")
  44         return np.concatenate([km, knm], axis=0)
  45 
  46     @property
  47     def labels(self) -> np.ndarray:

---

packages/scratch-core/src/conversion/plots/utils.py

  394 
  395 def _format_lr(llr_data: LLRData) -> str:
  396     """Format a single log-LR value with optional confidence interval."""
  397     if len(llr_data.llrs) != 1:
! 398         raise ValueError(f"expected single LR value, got {len(llr_data.llrs)}")
  399 
  400     log_lr = llr_data.llrs[0]
  401 
  402     if llr_data.llr_intervals is not None:
! 403         lower, upper = llr_data.llr_intervals[0][0], llr_data.llr_intervals[0][1]
! 404         return f"{log_lr:.2f} ({lower:.2f}, {upper:.2f})"
  405     return f"{log_lr:.2f}"
  406 
  407 
  408 def _common_results_metadata(

---

src/processors/controller.py

  47 
  48 
  49 def _extract_intervals(llr_data: LLRData) -> tuple[list[float], list[float]]:
  50     if llr_data.llr_intervals is None:
! 51         return [], []
  52     return llr_data.llr_intervals[:, 0].tolist(), llr_data.llr_intervals[:, 1].tolist()
  53 
  54 
  55 def compare_striation_marks(

---

src/processors/router.py

  64     include_in_schema=False,
  65 )
  66 async def calculate_score_impression(impression: CalculateScore) -> ComparisonResponseImpression:
  67     """Compare two impression profiles."""
! 68     vault = create_vault(impression.tag)
  69     return ComparisonResponseImpression.generate_urls(vault.access_url)
  70 
  71 
  72 @processors.post(

---

[github-actions]

Package	Line Rate	Branch Rate	Health
.	96%	88%	✔
comparators	100%	100%	✔
computations	100%	100%	✔
container_models	99%	100%	✔
conversion	97%	82%	✔
conversion.export	100%	100%	✔
conversion.filter	97%	89%	✔
conversion.leveling	100%	100%	✔
conversion.leveling.solver	100%	75%	✔
conversion.plots	98%	88%	✔
conversion.preprocess_impression	99%	91%	✔
conversion.preprocess_striation	89%	58%	✔
conversion.profile_correlator	96%	82%	✔
conversion.surface_comparison	93%	85%	✔
extractors	99%	75%	✔
mutations	100%	100%	✔
parsers	98%	80%	✔
parsers.patches	89%	60%	✔
preprocessors	99%	89%	✔
processors	98%	83%	✔
renders	98%	67%	✔
utils	92%	75%	✔
Summary	97% (3086 / 3174)	84% (331 / 396)	✔

Minimum allowed line rate is 50%

[SimoneAriens]

if this is what we want to return, why don't we just give back the whole reference_data? And I'm thinking we also want to return the interval of the new point, so the log_lr.

[SimoneAriens]

we need to talk about what we want to return and for what reason