Experiments included in the paper "Recursive maxima hunting: variable selection in functional data classification"

This repository contains the code for reproducing the experiments in the paper "Recursive maxima hunting: variable selection in functional data classification".

You will need an up-to-date installation of scikit-fda, numpy<2, setuptools<82 and scikit-datasets[utils-experiments] to run the code.

The executable files are:

• run_experiments.py: To run the real-data experiments.
• run_experiments_synthetic.py: To run the experiments with synthetic data.
• run_experiments_dependency_measure.py: To run the comparisons of different dependence measures.
• run_experiments_synthetic_time.py: To run the time comparisons.
• selection_example_all_datasets.py: Visualization of variables selected by RMH.

Apart from that, there are modules to download the results and plot them.