Re: [R] [R-sig-ME] Help with lmer
I´m not an expert either, but from my own experience with those annoying messages, I can suggest you some precautions you should take: - First of all, make sure, in your ".txt" file, that you are using dots for separating decimals, NOT commas. - Before writing your model, test each of the variables you are using (just type the name) and check that the software has not interpreted any covariate as a factor or viceversa. Hope that helps a bit. Good luck. De: Sophie Higgins Para: hugo.mildenber...@web.de; r-help@r-project.org; r-sig-mixed-mod...@r-project.org Enviado: dom,26 junio, 2011 13:26 Asunto: Re: [R-sig-ME] [R] Help with lmer Hello Hugo, Thank you for your reply. This is a snap shot of what my data looks like: Presencebsence Habitatype Substratetype Width Banktype BankIncline Bankheight Waterdepth 0 Lake Rocksgravel 600 Earth 0.45 less1 greater2 0 Lake Rocksgravel 600 Earth 0.45 less1 greater2 1 Lake Rocksgravel 600 Earth 0.45 less1 greater2 0 Lake Rocksgravel 600 Earth 0.45 less1 greater2 0 Lake Rocksgravel 600 Earth 0.45 less1 greater2 0 Stream Rocksgravel 1 Rocks 0.45 less1 025to05 0 Stream Rocksgravel 1 Rocks 0.45 less1 025to05 0 Stream Rocksgravel 1 Rocks 0.45 less1 025to05 0 Stream Rocksgravel 1 Rocks 0.45 less1 025to05 I am trying to find out if any off the above factors so substrate type, width, bank type etc have an effect on the presence/absence of the species in question. But as stated below I keep getting an error message up and I have no idea how to fix it in order for me to fun the lmer. Sophie > From: hugo.mildenber...@web.de > To: r-help@r-project.org > CC: england...@hotmail.com > Subject: Re: [R] Help with lmer > Date: Fri, 24 Jun 2011 17:19:00 +0200 > > Sophie, > > I'm myself quite new to linear mixed effect models, but "PresenceAbsence" >sounds like > a binary response variable. So wouldn't be > > m1<-lmer(Presencebsence~Habitatype*Width+(1|Sitename),familiy=binomial) > > the first thing to try? There is also a special mailing list for linear - >mixed -effect > > models: > > r-sig-mixed-mod...@r-project.org > > But with numeric problems like this one you really should provide a self >contained > > example (i.e. with data), at least the output of str(yourdata), if ever >possible. > > > Kind regards > > Hugo > > On Friday 24 June 2011 13:11:23 Sophie Higgins wrote: > > > > Hey, > > I am having trouble with lmer. I am looking at the presence/absence of > > water >shrews against habitat and other factors e.g > > so I used this: > > m1<-lmer(Presencebsence~Habitatype*Width+(1|Sitename))summary(m1) > > But i keep getting this error up > > Error in mer_finalize(ans) : Downdated X'X is not positive definite, 16.> >summary(m1)Error in asMethod(object) : matrix is not symmetric [1,2] > > What does this mean and now can I fix it? > > Sophie > > [[alternative HTML version deleted]] > > > > __ > > 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. [[alternative HTML version deleted]] ___ r-sig-mixed-mod...@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models [[alternative HTML version deleted]] __ 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] Plot confidence intervals
Hello, list: I'm not sure about where to send this question. I have several repeatability calculations, together with their 95% confidence intervals, and I would like to plot them, in a way similar to error bars. I was wondering if there is any specific function to do this, or any method I can apply in R, as I haven't found anything like that in the R book. Thank you very much Iker Vaquero-Alba Centre for Ecology and Conservation Daphne du Maurier Building University of Exeter, Cornwall Campus Treliever Road TR10 9EZ Penryn U.K. [[alternative HTML version deleted]] __ 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] no slot of name "fixef" for this object of class "lmerMod"
Hello everyone. I am trying to re-analyse some data with an R function I last used in 2011. Everything seemed to work fine then, but now, using the same code, it gives me this error: Error in R.pe(y, group1, group2, returnR = FALSE) : no slot of name "fixef" for this object of class "lmerMod" This is the part of the function that I think is relevant for the problem: # preparation #n <- rowSums(y) N <- length(y) k <- length(unique(group1)) # clone g <- length(unique(group2)) # round # require(lme4) # functions R.pe <- function(y, group1, group2, returnR=TRUE) { mod <- lmer(y ~ 1 + (1|group1)+(1|group2),verbose=FALSE) VarComp <- lme4::VarCorr(mod) beta0 <- as.numeric(mod@fixef) var.e <- attr(VarComp, "sc")^2 # residual variance var.a1 <- (as.numeric(VarComp[1])) # e.g. get clone R var.a2 <- (as.numeric(VarComp[2])) # e.g. get round R R.group1 <- var.a1/(var.a1+var.e) # clone level R.group2 <- var.a2/(var.a2+var.e) R.groupr <- var.a1/(var.a1+var.e+var.a2) if(returnR) return(list(R.group1=R.group1,R.group2=R.group2,R.groupr=R.groupr)) else return(list(beta0=beta0, var.e=var.e, var.a1=var.a1, var.a2=var.a2)) } I would appreciate any help about this. Let me know if you need more code, the function is obviously longer. Thank you very much in advance. Iker __________ Dr. Iker Vaquero-Alba Daphne du Maurier Centre for Ecology and Conservation College of Life and Environmental Sciences University of Exeter, Cornwall Campus TR10 9FE Penryn U.K. http://biosciences.exeter.ac.uk/cec/staff/postgradresearch/ikervaquero-alba/ https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] no slot of name "fixef" for this object of class "lmerMod"
Dear Thierry, Thank you very much. Do you mean I should add a line to the function with the code "fixef(my.model)", retaining everything else, even the "beta0 <- as.numeric(mod@fixef)" line? Or should I replace something? Also, when you say "fixef(my.model), do you mean "fixef(mod)" (as in my case "mod <- lmer(y ~ 1 + (1|group1)+(1|group2),verbose=FALSE)"? Thank you very much again. Iker ______ Dr. Iker Vaquero-Alba Daphne du Maurier Centre for Ecology and Conservation College of Life and Environmental Sciences University of Exeter, Cornwall Campus TR10 9FE Penryn U.K. http://biosciences.exeter.ac.uk/cec/staff/postgradresearch/ikervaquero-alba/ https://eric.exeter.ac.uk/repository/handle/10036/3381 De: Thierry Onkelinx Para: Iker Vaquero Alba CC: r-help@r-project.org Enviado: Viernes 3 de julio de 2015 11:00 Asunto: Re: [R] no slot of name "fixef" for this object of class "lmerMod" Dear Iker,The internals of the mer model have changed. Use fixef(your.model). Best regards, ir. Thierry Onkelinx Instituut voor natuur- en bosonderzoek / Research Institute for Nature and Forest team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance Kliniekstraat 25 1070 Anderlecht BelgiumTo call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of. ~ Sir Ronald Aylmer Fisher The plural of anecdote is not data. ~ Roger Brinner The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data. ~ John TukeyOp 2 jul. 2015 22:17 schreef "Iker Vaquero Alba" : Hello everyone. I am trying to re-analyse some data with an R function I last used in 2011. Everything seemed to work fine then, but now, using the same code, it gives me this error: Error in R.pe(y, group1, group2, returnR = FALSE) : no slot of name "fixef" for this object of class "lmerMod" This is the part of the function that I think is relevant for the problem: # preparation #n <- rowSums(y) N <- length(y) k <- length(unique(group1)) # clone g <- length(unique(group2)) # round # require(lme4) # functions R.pe <- function(y, group1, group2, returnR=TRUE) { mod <- lmer(y ~ 1 + (1|group1)+(1|group2),verbose=FALSE) VarComp <- lme4::VarCorr(mod) beta0 <- as.numeric(mod@fixef) var.e <- attr(VarComp, "sc")^2 # residual variance var.a1 <- (as.numeric(VarComp[1])) # e.g. get clone R var.a2 <- (as.numeric(VarComp[2])) # e.g. get round R R.group1 <- var.a1/(var.a1+var.e) # clone level R.group2 <- var.a2/(var.a2+var.e) R.groupr <- var.a1/(var.a1+var.e+var.a2) if(returnR) return(list(R.group1=R.group1,R.group2=R.group2,R.groupr=R.groupr)) else return(list(beta0=beta0, var.e=var.e, var.a1=var.a1, var.a2=var.a2)) } I would appreciate any help about this. Let me know if you need more code, the function is obviously longer. Thank you very much in advance. Iker __ Dr. Iker Vaquero-Alba Daphne du Maurier Centre for Ecology and Conservation College of Life and Environmental Sciences University of Exeter, Cornwall Campus TR10 9FE Penryn U.K. http://biosciences.exeter.ac.uk/cec/staff/postgradresearch/ikervaquero-alba/ https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] p-values in "rpt.adj"
Hello everyone, My question is about the "rptR" package. I am using the rpt.adj function (for adjusted repeatability). My model is as follows: rpt.adj(brifield ~ 1 + (1|ring)+(1|year),"ring",data=bellyfield,datatype="Gaussian") where "brifield" is the measurement for which I want to obtain repeatability and "ring" is the individual ID (there are 3 different measurements per ID – per ring number). What I want is an adjusted repeatability estimate that includes the year-to-year variance. Reading how to use the function, it says: grnameA character string or vector of character strings giving the name(s) of the grouping factor(s), for which the repeatability should be estimated. Spelling needs to match the random effect names as given in formula. So my grouping factor is "ring". Is that right? That would be my first question (just reassuring). And the second one is about the results: R = 0.591 SE = 0.148 CI = [0.27, 0.808] P = 1.06e-22 [] 0.001 [] It's basically about the p-values (as it gives me 2). Which one is which? The upper one is the p-value due to grouping factor "ring" (or ID) and the bottom one is the one of grouping factor "year"? Which one should I report, according to what I am interested in? Thank you very much in advance. Iker ______ Dr. Iker Vaquero-Alba Daphne du Maurier Centre for Ecology and Conservation College of Life and Environmental Sciences University of Exeter, Cornwall Campus TR10 9FE Penryn U.K. http://biosciences.exeter.ac.uk/cec/staff/postgradresearch/ikervaquero-alba/ https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] Potencial bug in R.adjust ("holm" method)
Hello all, I am doing some Bonferroni correction analyses with R.adjust function. I have a spreadsheet with 24 columns, each with 5 values. When I use the "holm" method, it gives me adjusted figures for all the original values except from the ones in the 4th row of each column. I mean, the value on the 4th row for every column is exactly the same either in the original data or in the corrected one. I've tried using another algorithm just to see what happens ("bonferroni", for example) and everything is fine, I get corrected figures for all the values, even the ones on the 4th row. Does anyone know whether this is any kind of known bug of the "holm" algorithm of P.adjust function. If so, should I worry about it? If so, can anybody suggest any possible solution? Thank you very much. Iker ______ Dr. Iker Vaquero-Alba Visiting Postdoctoral Research Associate Laboratory of Evolutionary Ecology of Adaptations, School of Life Sciences, University of Lincoln, Riseholme Park Campus, Lincoln LN2 2LG, UK. https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] Potencial bug in R.adjust ("holm" method)
Hello, Peter, and thank you for your clarifying reply. Actually, that was another doubt I had. As I take my data from different files (every "bunch of 5" values is taken from a different file), I assumed that, if I wanted to adjust for the error due to repeated measures, it should be enough to do it inside each file, right? I mean, the possible error due to repeated measures shouldn't go beyond each file. Once you close one file, open and attach a new one, the count, let's say, starts from scratch. Am I right? And if I'm not, what is the reason why should I pool all the p-values in a massive column and adjust them in bulk? Thank you very much for your help. Iker __________ Dr. Iker Vaquero-Alba Visiting Postdoctoral Research Associate Laboratory of Evolutionary Ecology of Adaptations, School of Life Sciences, University of Lincoln, Riseholme Park Campus, Lincoln LN2 2LG, UK. https://eric.exeter.ac.uk/repository/handle/10036/3381 De: peter dalgaard CC: "r-help@r-project.org" Enviado: Viernes 7 de agosto de 2015 14:40 Asunto: Re: [R] Potencial bug in R.adjust ("holm" method) > Hello all, > I am doing some Bonferroni correction analyses with R.adjust function. I >have a spreadsheet with 24 columns, each with 5 values. When I use the "holm" >method, it gives me adjusted figures for all the original values except from >the ones in the 4th row of each column. I mean, the value on the 4th row for >every column is exactly the same either in the original data or in the >corrected one. I've tried using another algorithm just to see what happens >("bonferroni", for example) and everything is fine, I get corrected figures >for all the values, even the ones on the 4th row. > > Does anyone know whether this is any kind of known bug of the "holm" >algorithm of P.adjust function. If so, should I worry about it? If so, can >anybody suggest any possible solution? > Thank you very much. > Iker p.adjust(), I presume? The Holm procedure is essentially to sort p-values in decreasing order and mutiplying by 1:n plus a little fiddling to keep the order and prevent p > 1. The logic is that if you reject the hypothesis corresponding to the smallest p after Bonferroni-correction by N, you only have N-1 simultaneous tests to consider, etc. The smallest multiplier will be 1, so it's not strange that one value appears uncorrected. It's curious that it is always in the 4th row, but it might be that the p-values in the 4 other rows are all considerably smaller. You do realize that if you run p.adjust for each column, you are not actually adjusting for the total of 120 tests, only for 5 of them, effectively ignoring the 23 other columns every time? -pd > > > __ > > Dr. Iker Vaquero-Alba > Visiting Postdoctoral Research Associate > Laboratory of Evolutionary Ecology of Adaptations, > School of Life Sciences, University of Lincoln, Riseholme Park Campus, >Lincoln > LN2 2LG, > UK. > > https://eric.exeter.ac.uk/repository/handle/10036/3381 > > > [[alternative HTML version deleted]] > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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. -- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd@cbs.dk Priv: pda...@gmail.com [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] Multinomial data
Hello everyone, I have some data I would like to analyse where my response variables are categorical data: several participants were asked to give answers in the form of 1 to 5 (for the degree of importance given to certain items). So, what I have now is response variables that can just take 5 possible values: 1, 2, 3, 4 or 5. After a bit of research, I have learned this might be a case of multinomial categorical data, although I am not sure whether ordered multinomial or nominal multinomial. I would like to apply generalized linear mixed models to my response variables, but I am a bit lost at the moment. Should I just use lmer and change the link function? Is it better to try MCMCglmm package and start from there? Any slight guidance or light shed on this would be greatly appreciated. Thanks a lot! __ Dr. Iker Vaquero-Alba Visiting Postdoctoral Research Associate Laboratory of Evolutionary Ecology of Adaptations School of Life Sciences Joseph Banks Laboratories University of Lincoln Brayford Campus, Lincoln LN6 7DL United Kingdom https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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] Factor response and explanatory variables.
Hello everyone, I am going to ask this certainly tricky question here not (yet) with the intention of getting a definitive answer, as I need to deepen my questions much more, but just to have an approximate idea of which direction taking next. I have a dataset where the potential response variables are categorical multinomial (ordered, I think): several people were asked to give a value from 1 to 5 to several attributes in a potential partner, both for a short term commitment or for a longer relationship. The age, religion, sexual orientation, sexual identity (gender), self-perceived sexual attractivess and minimum attractiveness demanded in a potential partner (this ones also with values 1 to 5) were recorded for each participant. The idea is using the values given to attributes in potential partners as the response variable, and see to what extent these are influenced by the person's age, gender, religion and so on. The problem is that all my variables are factors, I have no numeric ones. Also, as the values given to the same attributes for a short term commitment of for a longer relationship are expected to be correlated, I was considering using multiple response variables, which adds even more difficulty to the model. I have been reading about MCMCglmm, the course notes, which are not easy to understand. In any case, at some point I have read that if I don't want to fit random effects (as I think it's my case), I'd better use the pscl package instead. Can you give me any advice at this point? Should I try and use pscl, or is it better to try with MCMCglmm given the difficulty of the model? Any little help will be highly appreciated. Thank you very much Iker. ______ Dr. Iker Vaquero-Alba Visiting Postdoctoral Research Associate Laboratory of Evolutionary Ecology of Adaptations School of Life Sciences Joseph Banks Laboratories University of Lincoln Brayford Campus, Lincoln LN6 7DL United Kingdom https://eric.exeter.ac.uk/repository/handle/10036/3381 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.