Univariate analysis

• Compute logistic regression analysis (cross-validated prediction metrics and p-values) ~~~ ./code/univariate_logistic_regression.py ~~~
• Post-processing and data visualization ~~~ ./code/univariate.ipynb ~~~

Machine learning

• To launch one or all machine learning algorithms on various subsets of features (multivariate) ~~~ cd code python3 ml.py ~~~
• The parameters used (ml algorithms, selected features, etc…) have to be defined in ./code/params_ml.yaml

• The file ./code/requirements.txt contains versions of the python packages used

• Results are exported in output_path/feature_set/algorithm/ (see params_ml.yaml)

Post-processing

• Table summarizing some of the results. Parallel plot for performances and heatmap of misclassified patients ~~~ code/tables_results_ml_summary.ipynb ~~~