Wilhelm B. Kloke
2001-Aug-09 11:47 UTC
[R] converting BMDP 8V mixed model treatment to R ?
I am trying to translate a given BMDP 8V problem treatment to something approximately equivalent in nlme package (or base R, if sufficient). (Of course, it is not my goal to end there. I want to get access to the advanced flexibility of a more modern treatment.) Here is the problem statement in BMDP control language: /input title='Augenbewegungen'. variables=4. file='latm.dat'. format='11x,f7.0,f7.0,f12.4,f10.0'. /variables namesllat,rlat,vg,diff. /design ndep = 4. level=21,2,2,9,3. names = pers,retest,direct,flicker,ground. random = pers. fixed = retest,ground,direct,flicker. model = 'P,R,G,D,F'. /end. /finish. I tried in R (some minor differences in variable naming): latm <- read.table( "latm.dat" ) latm.aov <- aov( llat ~ direct * flicker *ground *retest +person + Error( person : ( direct * flicker *ground *retest ))) Indeed, summary(latm.aov) gives the same F tests than BMDP8V for the 1st part of the evaluation. But I don't get variance components from this. So it looks as if I have to change to mle from nmle package. I did not succeed to translate this model to mle yet. Trial-and-error fails mostly, because the problem thrashes my system (FreeBSD, K6-233, 64MB). This is particularly astonishing, as BMDP8V does it's work without problems on an old 24MB uVAX VMS system in mere 20 seconds CPU time. To run the above R partial solution I have to wait for about an hour on a freshly loaded R, after making the swapspace 3times as large as I use in normal work. There must be something wrong, but I have no idea, how to advance. -- Dipl.-Math. Wilhelm Bernhard Kloke Institut fuer Arbeitsphysiologie an der Universitaet Dortmund Ardeystrasse 67, D-44139 Dortmund, Tel. 0231-1084-257 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._