[R] Comparison between R and MATLAB

Earl F. Glynn efg at stowers-institute.org
Fri Jun 20 20:05:27 CEST 2008


"Shubha Vishwanath Karanth" <shubhak at ambaresearch.com> wrote in message 
news:7D44476A5833D340A0A85099D5EAD8B2832C at BAN-MAILSRV02.Amba.com...
Can I get a comparison between R and MATLAB? How is R efficient than MATLAB? 
Or what are the weaknesses of R compared to MATLAB?


Don't forget to compare licenses and cost.  Matlab's rigid and unreasonable 
license is the main reason I use R now.

Several years ago the Mathworks refused to let a post doc and me share a 
MatLab license -- they wanted all single-user licenses to be named users 
even if each of us only needed MatLab a few hours a month.  I asked if they 
had a "math model" for what we got for a license, but they didn't care that 
there was a huge disconnect between what they wanted to sell and our usage 
needs.  I quit using MatLab, converted my project from MatLab to R, and now 
steer as many people to R as possible.

The Mathworks refuses to acknowledge that life-science users of math tools 
are not like physical science users.  Biologists can spend weeks or months 
on experiments and then occasionally need high-end math tools  for analysis.

A single shared network MatLab license outside of academia is outrageously 
expensive.  A single network concurrent license for Matlab and the Image 
processing, Signal Processing and PDE toolboxes costs $18,800 -- I received 
that quote just yesterday since we still have a few people that use existing 
MatLab code.

MatLab may be a good product, but the Mathworks is unreasonable in how they 
license it.  We are a non-profit, basic life scientific research institute 
and the Mathworks sells us the same license as for-profit businesses.  Only 
degree-granting universities are eligible for academic pricing apparently.

efg
Earl F Glynn
Bioinformatics
Stowers Institute for Medical Research



More information about the R-help mailing list