Skip to content

Add documentation of applications.#54

Merged
GernotMaier merged 15 commits intoresiduals-overlearningfrom
res-learning-docu
Mar 25, 2026
Merged

Add documentation of applications.#54
GernotMaier merged 15 commits intoresiduals-overlearningfrom
res-learning-docu

Conversation

@GernotMaier
Copy link
Copy Markdown
Member

No description provided.

@GernotMaier GernotMaier self-assigned this Mar 23, 2026
@GernotMaier GernotMaier requested a review from Copilot March 25, 2026 18:58
@GernotMaier GernotMaier marked this pull request as ready for review March 25, 2026 18:59
Copy link
Copy Markdown
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR expands the stereo regression diagnostic surface area by caching additional training diagnostics in the model artifact and adding CLI tools to visualize/inspect those cached metrics, alongside README and changelog updates.

Changes:

  • Cache per-target generalization and residual-normality statistics during train_regression().
  • Add diagnostic_utils plus new diagnostic CLI scripts (generalization gap, PDP, permutation importance, residual normality, SHAP summary) and expose them via pyproject.toml console scripts.
  • Update README and towncrier fragment describing the new diagnostic capabilities.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 8 comments.

Show a summary per file
File Description
src/eventdisplay_ml/models.py Stores additional cached diagnostics (generalization_metrics, residual_normality_stats) during regression training.
src/eventdisplay_ml/diagnostic_utils.py New utility module for loading model artifacts, reconstructing splits, and computing cached diagnostic metrics.
src/eventdisplay_ml/scripts/diagnostic_generalization_gap.py New CLI to plot/train-test RMSE gaps (cache-first with fallback).
src/eventdisplay_ml/scripts/diagnostic_partial_dependence.py New CLI to compute and plot partial dependence diagnostics.
src/eventdisplay_ml/scripts/diagnostic_permutation_importance.py New CLI to compute permutation importance on reconstructed test split.
src/eventdisplay_ml/scripts/diagnostic_residual_normality.py New CLI to compute/plot residual normality diagnostics (cache-first with fallback).
src/eventdisplay_ml/scripts/diagnostic_shap_summary.py Minor logging/cleanup updates around cached SHAP plotting.
pyproject.toml Adds console entry points for the new diagnostic scripts.
tests/test_train_regression_standardization.py Extends regression training tests to validate cached generalization_metrics.
README.md Adds documentation for training/apply workflow and new diagnostic tools.
docs/changes/53.feature.md Updates changelog fragment describing new features/diagnostics.
.github/dependabot.yml Adds Dependabot configuration for GitHub Actions updates.

@GernotMaier GernotMaier merged commit c8fd277 into residuals-overlearning Mar 25, 2026
2 checks passed
@GernotMaier GernotMaier deleted the res-learning-docu branch March 25, 2026 19:17
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants