Message-ID: <CA+hbrhUTqstNPAqCZ+-hpgLeo+kmacadXCcOVMKVs3eppaw+PQ@mail.gmail.com>
Date: 2011-12-01T05:32:51Z
From: Peter Langfelder
Subject: Random Forests in R
In-Reply-To: <CAAyVsX+9TFwdUXZQndAmUD4ie=UHj9jVe+_ga-TgmM_TK5O54A@mail.gmail.com>
On Wed, Nov 30, 2011 at 7:48 PM, Axel Urbiz <axel.urbiz at gmail.com> wrote:
> I understand the original implementation of Random Forest was done in
> Fortran code. In the source files of the R implementation there is a note
> "C wrapper for random forests: ?get input from R and drive ?the Fortran
> routines.". I'm far from an expert on this...does that mean that the
> implementation in R is through calls to C functions only (not Fortran)?
>
> So, would knowing C be enough to understand this code, or Fortran is also
> necessary?
I haven't seen the C and Fortran code for Random Forest but I
understand the note to say that R code calls some C functions that
pre-process (possibly re-format etc) the data, then call the actual
Random Forest method that's written in Fortran, then possibly
post-process the output and return it to R. It would imply that to
understand the actual Random Forest code, you will have to read the
Fortran source code.
Best,
Peter