Hi, all. Are there any plans to spice up the principal components analysis tools in R? Many commercial packages offer interactive manipulation of 3D biplots. We are almost there already, with a fast (if homely-looking) biplot in "princomp." Any takers? BTW, BioConductor is fabulous. I went from downloading the packages to array analysis in three weeks. Thanks, everyone. Michael Benjamin, MD Emory Winship Cancer Institute Atlanta, GA [[alternate HTML version deleted]]

If you can get the SJava interface, you can do linked plots, brushing, and grand tours on scatterplots (i.e. principle components). The ROrca interface works pretty well these days, I just need more time to clean up the docs. best, -tony >>>>> "LG" == Laurent Gautier <laurent@cbs.dtu.dk> writes: LG> On Mon, Jan 27, 2003 at 11:29:39PM -0500, Michael Benjamin wrote: >> Hi, all. >> >> Are there any plans to spice up the principal components analysis tools >> in R? Many commercial packages offer interactive manipulation of 3D >> biplots. We are almost there already, with a fast (if homely- looking) >> biplot in "princomp." Any takers? LG> AFAIK, R can interface with ggobi. By linking selections of individuals LG> in different planes of projection, you should be able to achieve it LG> (and go beyond "only" 3 dimensions, if I understood what you meant LG> by "interactive manipulation"). LG> Check the screenshots at www.ggobi.org to get an idea. LG> Hopin' it helps, LG> L. LG> _______________________________________________ LG> Bioconductor mailing list LG> Bioconductor@stat.math.ethz.ch LG> http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor -- A.J. Rossini Rsrch. Asst. Prof. of Biostatistics U. of Washington Biostatistics rossini@u.washington.edu FHCRC/SCHARP/HIV Vaccine Trials Net rossini@scharp.org -------------- http://software.biostat.washington.edu/ ---------------- FHCRC: M: 206-667-7025 (fax=4812)|Voicemail is pretty sketchy/use Email UW: Th: 206-543-1044 (fax=3286)|Change last 4 digits of phone to FAX (my tuesday/wednesday/friday locations are completely unpredictable.)

On Mon, Jan 27, 2003 at 11:29:39PM -0500, Michael Benjamin wrote: > Hi, all. > > Are there any plans to spice up the principal components analysis tools > in R? Many commercial packages offer interactive manipulation of 3D > biplots. We are almost there already, with a fast (if homely- looking) > biplot in "princomp." Any takers? AFAIK, R can interface with ggobi. By linking selections of individuals in different planes of projection, you should be able to achieve it (and go beyond "only" 3 dimensions, if I understood what you meant by "interactive manipulation"). Check the screenshots at www.ggobi.org to get an idea. Hopin' it helps, L.

I think that we are likely to leave that to ggobi -- they have done a lot of stuff and the interface to R is pretty reasonable these days. On Tue, Jan 28, 2003 at 05:55:31AM +0100, Laurent Gautier wrote: > On Mon, Jan 27, 2003 at 11:29:39PM -0500, Michael Benjamin wrote: > > Hi, all. > > > > Are there any plans to spice up the principal components analysis tools > > in R? Many commercial packages offer interactive manipulation of 3D > > biplots. We are almost there already, with a fast (if homely- looking) > > biplot in "princomp." Any takers? > > AFAIK, R can interface with ggobi. By linking selections of individuals > in different planes of projection, you should be able to achieve it > (and go beyond "only" 3 dimensions, if I understood what you meant > by "interactive manipulation"). > Check the screenshots at www.ggobi.org to get an idea. > > > Hopin' it helps, > > > L. > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor -- +--------------------------------------------------------------------- ------+ | Robert Gentleman phone : (617) 632-5250 | | Associate Professor fax: (617) 632-2444 | | Department of Biostatistics office: M1B20 | Harvard School of Public Health email: rgentlem@jimmy.dfci.harvard.edu | +--------------------------------------------------------------------- ------+