01 Jan

More R Packages

This page has the following sections:

Using packages
Learning packages
Package quality
Crantastic
Major repositories

Using packages

In terms of you using a package, there are three states of a package:

  • The package exists somewhere.
  • The package is installed in your R library. This is often done with the install.packages function.
  • The package is attached in your session. This is often done with the require function.

Each of these states requires the preceding state.

Learning packages

There are two steps to being able to use a package. Step one is to find it. Suggestions on how to find the package are in More R search.

Once you’ve found a package, then you need to comprehend its functionality. If the package has a vignette, that is probably the best place to start.

You can see what vignettes are available to you with:

> vignette()

You can view a particular vignette with a command similar to:

> vignette("xts")

Package quality

The quality of the packages that ship with R is very high. The quality of contributed packages can run from abysmal to superb.

Most packages extend the functionality of R. There are some packages, though, that are meant to improve standard functionality of R.

A case of the latter is the plyr package. This package provides a more consistent set of tools for applying functions to pieces of datasets.

Crantastic

A resource that could be wonderful is Crantastic:

http://crantastic.org

This is a site that allows R users to rate and comment on packages in CRAN. This makes the quality of packages more visible. It is useful only if people rate and comment on packages. So: rate and comment on packages.

Major repositories

The major repositories for R packages are:

  • CRAN

The largest repository (by far) with over 4000 packages. This contains a staggering variety of packages.

main CRAN page (look at “Contributed extension packages”)

  • R-Forge

R-Forge is a platform where R packages are developed. Many packages have a stable version on CRAN and a development version on R-Forge. Some packages are still in the incubation stage and are only on R-Forge.

R-Forge

  • Bioconductor

Bioconductor consists of over 600 packages for analyses related to genomics.

Bioconductor

  • Omegahat

The Omega Project focuses on interfaces between R and other languages and systems.

omegahat.org

Back to top level of Impatient R

© Copyright - Burns Statistics