|
For statistical analysis, the most complete computing environment available is
the R programming language.
R started as GNU freeware alternative to the proprietary Splus.
In some senses, due to the massive contributions from the statistical
research community, it has become a better product than the latter.
My main software product is the ks (kernel smoothing)package
which implements a wide range of multivariate kernel smoothers (e.g. density estimation, density derivative estimation, classification, clustering, regression etc.), as well as sophisticated visualisation graphics. I started writing it as a doctoral candidate in the early 2000s, and I've carefully nurtured it
into its present form with a certain maturity.
This package is released under the General Public Licence (GPL >= 2) so anyone is free to download and use from the CRAN (Comprehensive R Archive Network), along with thousands of other user-contribted extension libraries to the base R distribution. It also forms the basis for the practical data analysis
exposited in the book
Multivariate Kernel Smoothing and Its Applications.
The statistical visualisations in the ks package are created in the base R graphics engine. These outputs are largely incompatible with the ggplot2 graphics engine, which forms the basis for statistical graphics in the tidyverse. To faciltate access for tidyverse users to the wide suite of kernel smoothers in ks, I have more recently developed the eks (extended kernel smoothing)package. This package is also compatible with geospatial data coded as simple features from the sf package.
Other sofware I've developed in R include:
A partial measure of the usage of the CRAN packages can be gleaned from the download statistics from the RStudio CRAN mirror:
|
Current version |
Last 30 days |
Since Oct 2012 |
Age |
| ks |
 |
 |
 |
 |
| eks |
 |
 |
 |
 |
| feature |
 |
 |
 |
 |
| prim |
 |
 |
 |
 |
|