rpf_benchmark

Benchmarking randomPlantedForest for classification against ranger and xgboost.

Setup

• batchmark.R contains learner definitions with parameter spaces and batchtools setup via mlr3batchmark.
• analysis.R is used to investigate results.
• get_oml_tasks.R is used to retrieve a benchmarking suite from OpenML.

The project uses renv to track dependencies, which is particularly important for GitHub dependencies (randomPlantedForest, mlr3batchmark) and our fork of mlr3extralearners which contains the experimental classif.rpf learner used here (see here).

OpenML usage benefits from setting an API key, see here.