[Feature Request] proposal of add an option choosing random forests implementation (Breiman or ranger)

[armgong]

Ranger is a fast implementation of random forests,its fast than the Breiman version. could we add an option on the randomForest related classfication,clustering,regression?
IMO, we can add an option and the "algorithm settings" tab, add it before "Training data used per tree" ,allow user choice use the Breiman or Ranger implementation of random forests

btw, I read the code and add some codes to do this , but no clues how to test the modified jaspmachinelearing module (don't want compile the whole jasp).
if someone could give some hints how to, I will submit a patch about this proposal.

[shun2wang]

Hi @armgong Thanks for your contribution. Yes compiling JASP is not necessary. You can install your change and test it by following here: https://github.com/jasp-stats/jasp-desktop/blob/development/Docs/development/jasp-module-workflow.md

For more development information you can read our dev docs. feel free to ask question here.