### Getting started with R statistical language

R is a programming language and software environment for statistical computing and graphics. The R language is widely used among statisticians for developing statistical software, and R is widely used for statistical software development and data analysis.

Recently, I started reading about R. I found following resources mentioned in references section to learn R. I was particularity interested in R with GPU and R with Hadoop. I will update this post again once I am ready with some good examples of R + GPU and R + Hadoop. It would be even better if I can make an example use case with R + Hadoop + GPU.

R with GPU computing:

Recent advances in consumer computer hardware makes parallel computing capability widely available to most users. Applications that make effective use of the so-called graphics processing units (GPU) have reported significant performance gains.

One of the most affordable GPU available is NVIDIA’s CUDA. Incidentally, the CUDA programming interface is vector oriented, and fits perfectly with the R language paradigm.

R will not deal with CUDA directly or its advanced C/C++ interface. Instead, R will rely on rpud and other R packages for studying GPU computing. For more details look at link in references section.

As it turns out, I need to have CUDA compatible GPU to test R + CUDA samples. Without it I will not be able to run R + CUDA samples. CUDA does not have a CPU fallback if comptible GPU is not found. Latest CUDA toolkit (see the CUDA toolkit link in references section) does not event install. This is totally unlike AMD Aparapi which falls back to CPU if compatible GPU is not found on host machine.

References:

Recently, I started reading about R. I found following resources mentioned in references section to learn R. I was particularity interested in R with GPU and R with Hadoop. I will update this post again once I am ready with some good examples of R + GPU and R + Hadoop. It would be even better if I can make an example use case with R + Hadoop + GPU.

R with GPU computing:

Recent advances in consumer computer hardware makes parallel computing capability widely available to most users. Applications that make effective use of the so-called graphics processing units (GPU) have reported significant performance gains.

One of the most affordable GPU available is NVIDIA’s CUDA. Incidentally, the CUDA programming interface is vector oriented, and fits perfectly with the R language paradigm.

R will not deal with CUDA directly or its advanced C/C++ interface. Instead, R will rely on rpud and other R packages for studying GPU computing. For more details look at link in references section.

As it turns out, I need to have CUDA compatible GPU to test R + CUDA samples. Without it I will not be able to run R + CUDA samples. CUDA does not have a CPU fallback if comptible GPU is not found. Latest CUDA toolkit (see the CUDA toolkit link in references section) does not event install. This is totally unlike AMD Aparapi which falls back to CPU if compatible GPU is not found on host machine.

References:

- WIKI: http://en.wikipedia.org/wiki/R_%28programming_language%29
- R introduction: http://www.r-tutor.com/r-introduction
- GPU Computing with R: http://www.r-tutor.com/gpu-computing
- R + CUDA Toolkit Download: http://www.r-tutor.com/content/download
- Latest version of CUDA toolkit download link: http://developer.nvidia.com/cuda-toolkit-40
- Free R eBooks: http://stackoverflow.com/questions/192369/books-for-learning-the-r-language
- Other free statistics eBooks: http://www.r-statistics.com/2009/10/free-statistics-e-books-for-download/
- R Notes: http://www.johndcook.com/R_language_for_programmers.html

## Comments

## Post a Comment