BACKGROUND

"Random Jungle" is a fast implementation of the Random Forests algorithm, developed
and released under an open source license by D. Schwarz, I. Konig, and A. Ziegler [1].
See http://randomjungle.sourceforge.net/ for information about the Random Jungle project.

RandomJungle::Jungle provides a simplified interface to access Random Jungle
input and output data. See RandomJungle::Tree and RandomJungle::Tree::Node for
methods relating to the classification trees produced by Random Jungle.  The other modules
in the RandomJungle bundle (e.g., RandomJungle::File::*) are primarily for internal use.

SCOPE

The RandomJungle* modules in this release DO NOT implement the Random Forests algorithm.
Instead, they are designed to provide a simplified interface to the input and output data
files for the Random Jungle implementation.

Furthermore, these modules are not intended to be a comprehensive library.  They were created
to solve specific problems encountered by a research team.  Specifically, these modules were
designed to work under the parameters in which our team uses Random Jungle (e.g., genotype
data in a specified ped format, out-of-bag and XML output file parameters set).  An example
of how we call Random Jungle is as follows:

rjunglesparse-1.0.361_static --file=input.raw --ntree=100 --depvarname=PHENOTYPE
--outprefix=outputfile --impmeasure=3 --memmode=2 --maxtreedepth=100
--targetpartitionsize=1 --pedfile -w2 --oob

See the files that are included in the /t directory for examples of file format.  See also the
/scripts directory for examples of how these modules can be used with the input/output files.

These modules will not provide full functionality if other command line options (e.g., file formats)
are used.  In that case, these modules will need to be extended (I tried to design them with that
in mind).  Code contributions are welcome.

INSTALLATION

You should be able to use this set of instructions to install the module:

perl Makefile.PL
make
make test
make install

If you are on a windows box you should use 'nmake' rather than 'make'.

REFERENCES

[1]:  Schwarz et al.  On safari to Random Jungle: a fast implementation of Random Forests for high-dimensional data.
Bioinformatics 26(14):1752-8. July 2010. http://www.ncbi.nlm.nih.gov/pubmed/20505004