[R] GLMER Syntax Question
I was wondering is someone can explain me the differences between these (random slopes and intercept) models model1 - glmer(output~A+B+C+(A+B+C | F) ) model2 - glmer(output~A+B+C+(A | F)+(B | F)+(C | F) Thanks. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] library REEMtree = Error in estRE[toString(uniqueID[i]), 1] : incorrect number of dimensions
Hi List, I'm having this problem when trying to use the PREDICT function. Here is a way to reproduce the error library(REEMtree) data(simpleREEMdata) REEMresult-REEMtree(Y~D+t+X, data=simpleREEMdata, random=~1|ID/D) predict(REEMresult, simpleREEMdata, id = simpleREEMdata$ID/simpleREEMdata$D, EstimateRandomEffects=TRUE) As far as I understand the problem is that I'm not being able to specify correctly the number of groups created in the random structure of the model (~1|ID/D). Does anybody knows how to FIX this ? __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] is there a way/library for generating colorful noise in R ??
I would like to generate some noisy time series. I know that it is possible to classify noise by looking at the exponent (beta) of the relationship between the spectrum of the time series and the frequencies (i.e. spectrum ~ frequency ^ beta ). Is there a way to generate White (beta=0), Pink (beta=-1), Brown (Beta=-2), Blue(beta=1) and Violet (beta=2) noise in R ?. Thanks. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] is this an ANOVA ?
Hi all, I have a very easy questions (I hope). I had measure a property of plants, growing in three different substrates (A, B and C). The rest of the conditions remained constant. There was very high variation on the results. I want to do address, whether there is any difference in the response (my measurement) from substrate to substrate? x-c('A','A','A','A','A','B','B','B','B','B','C','C','C','C','C') # Substrate type y - c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15) # Results of the measurement MD-data.frame(x,y) I wrote a linear model for this: summary(lm(y~x,data=MD)) This is the output: Call: lm(formula = y ~ x, data = MD) Residuals: Min 1Q Median 3QMax -2.000e+00 -1.000e+00 5.551e-17 1.000e+00 2.000e+00 Coefficients: Estimate Std. Error t value Pr(|t|) (Intercept) 3. 0.7071 4.243 0.001142 ** xB5. 1. 5.000 0.000309 *** xC 10. 1. 10.000 3.58e-07 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 Residual standard error: 1.581 on 12 degrees of freedom Multiple R-squared: 0.8929, Adjusted R-squared: 0.875 F-statistic:50 on 2 and 12 DF, p-value: 1.513e-06 I conclude that there is an effect of substrate type (x). NOW the questions : 1) Do the fact that the all p-values are significant means that all the groups are different from each other ? 2) Is there a (easy) way to plot, mean plus/minus 2*sd for each substrate type ? (with asterisks denoting significant differences ?) THANKS ! version platform x86_64-apple-darwin9.8.0 arch x86_64 os darwin9.8.0 system x86_64, darwin9.8.0 status major 2 minor 11.1 year 2010 month 05 day31 svn rev52157 language R version.string R version 2.11.1 (2010-05-31) __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] LMER: How to specify Random Effects
I saw different specifications for Random Effects and I'm confused about the use of / and the use of (0+...|) . Let say we have a nested structure where some countries have some several plants in different states and we measure the reaction to a drug. The list of Countries = USA, France, Italy The States for USA = Michigan, Florida, California The States for France = Paris, Orleans The States for Italy = Venezia, Sienna, Florence, Rome, Napoli , Sicilia Plants were classified as High and Low is this the way to specify a possible model ? lmer(Reaction ~ Drug + (1| Country / State / Plant) , data) or should I use something like this A) lmer(Reaction ~ Drug + (0| Country / State / Plant) , data) B) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State / Plant) , data) C) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State / Plant) + (0+Country + State | Plant), data) D) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State ) + (0+Country + State | Plant), data) Thanks __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] LMER: How to specify Random Effects
Thanks for your feedback. Actually Plant is nested since High and Low are qualitative relative values (High in Michigan is not the same as High in Sienna). S Ellison wrote: The Plant classification is not nested; it's an effect across all countires and states and probably a fixed effct (assuming you want to measure its size or significance). But the state is nested in country. That would suggest to me lmer(Reaction~Drug+Plant+(1|Country/State),...) (or Plant*Drug, if you want an interaction) However, though your states are nested in countrythey are readily identifiable by lmer as different for each country (you do not have country1: State1, State2..., Country2:State1,State2..) so the nesting by country probably doesn't need to be specified. I think the above would, _in this case_ be almost identical to lmer(Reaction~Drug+Plant+(1|Country)+(1|State),...) Ubuntu Diego ubuntu.di...@gmail.com 01/12/2009 16:22:09 I saw different specifications for Random Effects and I'm confused about the use of / and the use of (0+...|) . Let say we have a nested structure where some countries have some several plants in different states and we measure the reaction to a drug. The list of Countries = USA, France, Italy The States for USA = Michigan, Florida, California The States for France = Paris, Orleans The States for Italy = Venezia, Sienna, Florence, Rome, Napoli , Sicilia Plants were classified as High and Low is this the way to specify a possible model ? lmer(Reaction ~ Drug + (1| Country / State / Plant) , data) or should I use something like this A) lmer(Reaction ~ Drug + (0| Country / State / Plant) , data) B) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State / Plant) , data) C) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State / Plant) + (0+Country + State | Plant), data) D) lmer(Reaction ~ Drug + (1| Country ) + (0+Country | State ) + (0+Country + State | Plant), data) Thanks __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. *** This email and any attachments are confidential. Any u...{{dropped:9}} __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] SNOW: Error in socketSelect(socklist) : not a socket connection
I'm trying to use snow in my dual-core (hopefully later this is going to run in a cluster). So, at this moment I create a cluster using SOCK connection (MPI in the future). However when I try to use clusterApplyLB I got Error in socketSelect(socklist) : not a socket connection. Any ideas ? Do you know if that is going to be an isuue too when I swith from SOCK to MPI ? Sample code is attached. Thanks. # TODO: Add comment # # Author: diego ### rm(list=ls()) library(snow) t1-Sys.time() rates-c(0.5,0.5,0.7) initialState-c(0,0,0,0,0) AllEvents-c(1,1,1,1,2,2,2,3,1,2) AllLocations-c(1,2,3,4,1,4,2,1,5,4) AllTimes-1:10 AllConfigurations-rbind( c(0,0,0,0,0), c(1,0,0,0,0), c(1,1,0,0,0), c(1,1,1,0,0), c(1,1,1,1,0), c(2,1,1,1,0), c(2,1,1,2,0), c(2,2,1,2,0), c(0,2,1,2,0), c(0,2,1,2,1), c(0,2,1,0,1) ) nevents-length(AllEvents) clusterSize-2 cl-makeCluster(clusterSize,type=SOCK) arguments-vector(list,nevents) completedTimes-c(0,AllTimes) for(i in 1:nevents){ element-c(AllConfigurations[i,], rates, completedTimes[i], AllEvents[i], AllLocations[i], AllTimes[i]) arguments[[i]]-element } source(GetLogLikelihood.R) parallelOutputs-clusterApplyLB(cl,arguments,GetLogLikelihood) # This give me error #parallelOutputs-clusterApply(cl,arguments,GetLogLikelihood) # This Work OK print(sum(unlist(parallelOutputs))) stopCluster(cl) t2-Sys.time() print(t2-t1) __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Snow Parallel R: makeCluster with more nodes than available
Hi all, I would like to know what would happen if using snow I create a cluster of size 50, for example using makeCluster(50,type='SOCK') on a machine with 2 Cores and run a function. Does snow run 25 and 25 functions on each of my 2 real processors or it just run 50 functions in one processor ? Thanks. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Memory issue?
I had similar issues with memory occupancy. You should explicitly call gc() to call the garbage collector (free memory routine) after you do rm() of the big objects. D. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.