feat: record training history

[JemmaLDaniel]

Record training history during calibrator fitting

Summary

Adds a TrainingHistory dataclass and integrates it into the calibrator's fit method so that training metrics (loss curve, validation scores, iteration count) are captured and persisted after training. This enables post-hoc analysis of training convergence and makes it easier to diagnose issues like underfitting or early stopping behaviour.

Changes

• winnow/calibration/calibrator.py:
  • Adds TrainingHistory dataclass with fields: loss_curve, validation_scores, final_training_loss, final_validation_score, n_iter.
  • TrainingHistory.save(path) / TrainingHistory.load(path) — JSON serialisation for persisting histories.
  • TrainingHistory.plot(output_path, show) — matplotlib visualisation of loss and validation curves.
  • Calibrator.fit() now returns a TrainingHistory instance extracted from the fitted MLPClassifier.
• winnow/calibration/__init__.py — re-exports TrainingHistory.
• winnow/scripts/main.py — the train CLI command now saves the training history JSON alongside the model output.
• winnow/configs/train.yaml — adds history_output config key for specifying the history file path.
• docs/api/calibration.md — documents TrainingHistory and its methods.
• tests/calibration/test_calibrator.py — tests that fit returns a TrainingHistory with expected fields, and tests for save/load/plot round-tripping.

[github-actions]

Coverage Report

File	Stmts	Miss	Cover	Missing
__init__.py	0	0	100%
data_types.py	4	0	100%
calibration
__init__.py	2	0	100%
calibration_features.py	316	7	97%	247–248, 443, 728, 916, 920, 1215
calibrator.py	150	16	89%	158, 212–213, 215, 249–252, 277–278, 280, 305–306, 310, 337–338
compat
__init__.py	0	0	100%
instanovo.py	10	6	40%	12, 14–15, 17, 24–25
datasets
__init__.py	0	0	100%
calibration_dataset.py	109	17	84%	155, 169, 171, 173, 183, 196, 249, 251–252, 258–261, 263–266
data_loaders.py	270	14	94%	23, 189, 220–221, 414, 455, 847, 851, 900, 911, 1023–1024, 1052–1053
interfaces.py	3	0	100%
psm_dataset.py	25	0	100%
fdr
__init__.py	0	0	100%
base.py	58	15	74%	81, 85–86, 91, 98–99, 105, 126, 129–130, 135, 137–138, 144, 186
database_grounded.py	28	1	96%	52
nonparametric.py	25	4	84%	62, 68–69, 72
scripts
__init__.py	0	0	100%
main.py	191	191	0%	8, 10–13, 16–20, 23–24, 26–28, 32, 39, 44, 47, 53, 55–56, 59, 68, 76, 79, 86, 88–90, 92, 94–99, 102, 104–105, 110, 125, 128, 135–141, 144–145, 148, 161–163, 166, 169, 174, 176–178, 180, 182–183, 186–187, 190, 192–193, 195, 197, 199–200, 202, 205–206, 209–210, 213–216, 221–222, 225–226, 229–231, 233, 236, 250–252, 254, 256, 261, 263–265, 267–268, 270–272, 277–278, 280–282, 284, 286, 288–289, 293–296, 298–299, 301–302, 304–305, 307, 310, 324–326, 329, 332, 337, 339–341, 343–345, 347–348, 351–352, 355, 357–358, 360, 362, 364–365, 367, 370–371, 377–378, 381–382, 385–386, 389–390, 398–400, 404, 407, 411, 414, 437, 450–451, 454, 476, 488–489, 492, 517, 530–531, 534, 549, 561–562, 565, 577, 589–590, 593, 608, 620–621
utils
__init__.py	4	0	100%
config_formatter.py	53	40	24%	29, 37–38, 40–42, 44, 55, 58–60, 62–63, 66–69, 72–74, 77–78, 80, 91, 102, 113, 127–128, 130–132, 145–147, 150, 153–154, 157–158, 160
config_path.py	76	5	93%	24–26, 117–118
peptide.py	16	0	100%
TOTAL	1340	316	76%

Tests	Skipped	Failures	Errors	Time
299	0 💤	0 ❌	0 🔥	36.751s ⏱️

[JemmaLDaniel]

This PR is encompassed in #190