[R-sig-ME] lme4 missing from repositories?

Federico Calboli f.calboli at imperial.ac.uk
Thu Oct 21 11:22:14 CEST 2010

On 21 Oct 2010, at 03:32, huang min wrote:

> For your codes, R32 is about 20% faster on my machine and as Simon Urbanek suggested, R32 is more well tested. Though I forgot when he mentioned this (Maybe a bit outdated).
> I don't know whether there are recently information about the 32bit vs 64bit.

I think it is reasonable to make the case that, for the use of the library lme4, in order to get results you are happy to send to a journal you have to use R64. The 20% time increase is nothing compared to letters of retraction, unpublished papers because you are not sure of the results, and possibly a license for a different software to doublecheck the results.

Additionally 64 bit architecture is not exactly new stuff now, and I've been running R64 on Mac and Linux for years. I'm surprised to hear that R64 is not as well tested as R32 (though I am surprised to hear that R32 is faster for the example I posted).



Federico C. F. Calboli
Department of Epidemiology and Biostatistics
Imperial College, St. Mary's Campus
Norfolk Place, London W2 1PG

Tel +44 (0)20 75941602   Fax +44 (0)20 75943193

f.calboli [.a.t] imperial.ac.uk
f.calboli [.a.t] gmail.com

More information about the R-sig-mixed-models mailing list