Weighted Subspace Random Forest for Classification

A parallel implementation of Weighted Subspace Random Forest. The Weighted Subspace Random Forest algorithm was proposed in the International Journal of Data Warehousing and Mining by Baoxun Xu, Joshua Zhexue Huang, Graham Williams, Qiang Wang, and Yunming Ye (2012) . The algorithm can classify very high-dimensional data with random forests built using small subspaces. A novel variable weighting method is used for variable subspace selection in place of the traditional random variable sampling.This new approach is particularly useful in building models from high-dimensional data.

License Version on CRAN Number of downloads from RStudio CRAN mirror

The wsrf is a parallel implementation of the Weighted Subspace Random Forest algorithm (wsrf) of Xu et al. A novel variable weighting method is used for variable subspace selection in place of the traditional approach of random variable sampling. This new approach is particularly useful in building models for high dimensional data---often consisting of thousands of variables. Parallel computation is used to take advantage of multi-core machines and clusters of machines to build random forest models from high dimensional data with reduced elapsed times.

Documentation & Examples

The package ships with a html vignette including more details and a few examples.


Currently, wsrf requires R (>= 3.3.0), Rcpp (>= 0.10.2). For the use of multi-threading, a C++ compiler with C++11 standard support of threads (for example, GCC 4.8.1) is required. Since the latest version of R has added support for C++11 on all operating systems, we do not provide support for the old version of R and C++ compiler without C++11 support. To install the latest version of the package, from within R run:

R> install.packages("wsrf")


Previous version of wsrf provide support on systems without C++11 or using Boost for multithreading. Though we do not provide support for these options anymore, but still leave the usage here for someone with needs of previous version of wsrf. The choice is available at installation time depending on what is available to the user:

# To install previous version of wsrf without C++11
R> install.packages("wsrf", type = "source", configure.args = "--enable-c11=no")
# To install previous version of wsrf with Boost for multithreading
R> install.packages("wsrf",
+                   type = "source",
+                   configure.args = "--with-boost-include=<Boost include path>
                                      --with-boost-lib=<Boost lib path>")

After installation, one can use the built-in function wsrfParallelInfo to query whether the version installed is what they really want (distributed or multi-threaded).

R> wsrfParallelInfo()


GPL (>= 2)


Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.7.17 by He Zhao, a year ago

https://github.com/SimonYansenZhao/wsrf, http://togaware.com

Report a bug at https://github.com/SimonYansenZhao/wsrf/issues

Browse source code at https://github.com/cran/wsrf

Authors: Qinghan Meng [aut] , He Zhao [aut, cre] , Graham J. Williams [aut] , Junchao Lv [aut] , Baoxun Xu [aut] , Joshua Zhexue Huang [aut]

Documentation:   PDF Manual  

Task views: Machine Learning & Statistical Learning

GPL (>= 2) license

Depends on parallel, Rcpp, stats

Suggests knitr, party, randomForest, rattle.data, stringr

Linking to Rcpp

System requirements: C++11

See at CRAN