Marvin N. Wright

Implements random forests (RF, Breiman 2001) with emphasis on simplicity. Uses reference classes and only plain R. Not optimized for computation speed. Allows rapid prototyping of RF-type algorithms.

For a runtime-optimized random forest implementation, see ranger.

Features:

• Variable importance
• Use a standard predict() function
• ...

Technical stuff:

• Software tests
• Profile the code. Possible to make it faster without losing simplicity?

Be careful, this package is not extensively tested!