R help - Jul 2006 - Reproducible Research - Examples

All,

Recently I ran across a URL documenting published research using R: 

http://www.cgd.ucar.edu/ccr/ammann/millennium/CODES_MBH.html

A note on the site indicates that the code is being revised. The code and
data are provided, so that one could reproduce the results without having to
buy a proprietary software program. In poking around the R website it is
clear that a lot of thought has gone into documenting reproducible research,
notably by Harrell, Gentleman, and the Sweave effort, among others.

http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/StatReport
http://cran.ssds.ucdavis.edu/doc/contrib/Harrell-statcomp-notes.pdf
http://www.bioconductor.org/docs/papers/2003/Compendium

Question: Reproducible research is clearly desirable and feasible. Could
anyone provide examples (stand-alone URLs or supplementary material in
journals) that you would recommend as a models for reproducible research in
R?

Thanks,

Phil

Philip Heilman
USDA-ARS Southwest Watershed Research Center
2000 E. Allen Rd.
Tucson, AZ 85704
(520) 670-6380 x148
pheilman at tucson.ars.ag.gov
http://www.tucson.ars.ag.gov/

>  Question: Reproducible research is clearly desirable and feasible. Could
>  anyone provide examples (stand-alone URLs or supplementary material in
>  journals) that you would recommend as a models for reproducible research
in
>  R?
Here are some attempts based on RWeb, contributed by Jeff Banfield, to allow
for the reproduction of research papers:

http://biomserv.univ-lyon1.fr/~necsulea/repro/
http://pbil.univ-lyon1.fr/members/lobry/repro/gene06/
http://biomserv.univ-lyon1.fr/~palmeira/repro/
http://pbil.univ-lyon1.fr/members/lobry/repro/toxo/
http://pbil.univ-lyon1.fr/datasets/SemonLobryDuret2005/
http://pbil.univ-lyon1.fr/members/lobry/repro/gene05/
http://pbil.univ-lyon1.fr/members/lobry/repro/bioinfo04/
http://pbil.univ-lyon1.fr/members/lobry/repro/lncs04/
http://pbil.univ-lyon1.fr/members/lobry/repro/jag03/
http://pbil.univ-lyon1.fr/members/lobry/repro/jme97/
http://pbil.univ-lyon1.fr/members/lobry/repro/jme95/
http://pbil.univ-lyon1.fr/members/lobry/repro/cabios95/
http://pbil.univ-lyon1.fr/members/lobry/repro/nar94/

I would not recommend these as a "models", but I would like to point
out that:

1) If you are already using Sweave (many thanks to Friedrich Leisch) there
    is virtually no extra-cost for the authors of the paper to write the
    corresponding web page.

2) There is no way to be more cristal clear with respect to your reviewers
    during the submission process: data are available online, methods are
    available online, the parameters of the methods are available and
    editable online.

This is good science, but, problem: what would be the equivalent
of JSTOR for this? How people can, say 100 years in the future,
reproduce your results?

Best,

Jean Lobry


P.S. If you want to play with the data used by Student (aka William Sealy
Gosset) for the experimental validation of the famous t-test
(The probable error of a mean. Biometrika, 6:1-25,1908) they are at
your R prompt:
crim<-read.table(file="http://pbil.univ-lyon1.fr/R/donnees/criminals1902.txt")
Can you reproduce Student's results nowadays?

-- 
Jean R. Lobry            (lobry at biomserv.univ-lyon1.fr)
Laboratoire BBE-CNRS-UMR-5558, Univ. C. Bernard - LYON I,
43 Bd 11/11/1918, F-69622 VILLEURBANNE CEDEX, FRANCE
allo  : +33 472 43 12 87     fax    : +33 472 43 13 88
http://pbil.univ-lyon1.fr/members/lobry/