oadr_cpep package

Submodules

• oadr_cpep.aggregate module
  • consensus_features()
  • aggregate_vectors()
• oadr_cpep.apply module
  • apply_coefficients()
• oadr_cpep.cli module
  • select_features_command()
  • fit_ridge_command()
  • fit_lasso_command()
  • fit_rf_command()
  • fit_models_command()
  • apply_coefficients_command()
  • consensus_features_command()
  • aggregate_vectors_command()
  • main()
• oadr_cpep.common_utils module
  • load_site()
  • read_feature_list()
  • stem()
  • design_matrix()
  • kfold()
  • cv_predict()
  • r2()
  • mse()
  • bootstrap_r2_ci()
• oadr_cpep.fit module
  • fit_ridge()
  • fit_lasso()
  • fit_rf()
  • fit_models()
• oadr_cpep.logging_config module
  • setup_logger()
• oadr_cpep.oadr_data module
  • load_panel_a()
  • load_panel_b()
  • panel_b_design_matrix()
  • panel_a_design_matrix()
  • load_features()
• oadr_cpep.plot module
  • scatter()
  • solo_vs_federated()
• oadr_cpep.select module
  • select_features()

Module contents

oadr-cpep: Federated prediction of residual beta-cell function (C-peptide AUC) in Type 1 Diabetes.

One typer CLI; one function per process, grouped into logical modules:

select.py : select_features (Phase 1) fit.py : fit_ridge, fit_lasso, fit_rf, fit_models (Phase 2) apply.py : apply_coefficients (Phase 3, site outcome) aggregate.py : consensus_features, aggregate_vectors (aggregator) plot.py : all graphics (matplotlib PNG/SVG, plotly HTML) common_utils.py, oadr_data.py, logging_config.py (shared)

Only model parameters and scalar performance summaries cross the site boundary — never subject-level data.