[R-sig-ME] estimated growth curves from lme4

Sun Jul 15 13:54:22 CEST 2007

On 7/14/07, Afshartous, David <afshart at exchange.sba.miami.edu> wrote:
> All,
>
> The archive email below mentions the lack of general predict methods in
> lme4.  I would like to plot the estimated growth curves for
> treatment/placebo
> groups (or the estimated patient growth curves that augment the former
> w/ the patient estimated
> random effects).
>
> I obtained the plot manually by constructing the appropriate functions
> from the fixed effects output and plotting using curve() (sample code
> below).
>
> For those that have transitioned from lme, do you suggest sticking w/
> such an approach
> (and possibly automating it w/ additional code, but this would be
> dependent on the type
> of model that is estimated each time), or going back to lme where I
> recall that this was relatively
> simple given the nesting mentioned by Douglas Bates below.
>
> Thanks,
> David
>
> sample code:
> ## assume growth curve is quadratic, and coefficeints for placego group
> are 10, 50, -20:
> quad.fun.Placebo <- function(t) {
> y = 100 + 50*t - 20*t^2
> y}
> ## plot from t=0 to t=5
> curve(quad.fun.Placebo, 0,5, xlab="Time", ylab="Dependent Variable",
> col="red")
>
> ps - why don't the methods show up below?
> > library(lme4)
> Loading required package: Matrix
> Loading required package: lattice
> > methods(class="lme4")
> no methods were found

I think you meant class = "lmer" (there is no class called "lme4").
Even that won't show any methods, however, because the lme4 package
uses S4 classes and methods, which is the reason for the "4" in the
name of the package.  The methods function only shows S3 methods.  To
see the listing of methods defined in the lme4 package use

library(lme4)
showMethods(where = "package:lme4")

> From: Douglas Bates <bates_at_stat.wisc.edu
> <mailto:bates_at_stat.wisc.edu?Subject=Re:%20%5BR%5D%20Predicted%20value
> s%20in%20lmer%20modeling> >
> Date: Tue 28 Nov 2006 - 13:53:23 GMT
>
>
> On 11/28/06, Fucikova, Eva <E.Fucikova at nioo.knaw.nl> wrote:
> > Dear All,
>
>
> > I am working with linear mixed-effects models using the lme4 package
> in
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html#6180qlink1>
> > R. I created a model with the lmer function including some main
> effects,
> > a two-way interaction and a random effect. Now I am searching for a
> way
> > to save the predicted values for this model.
>
>
> > As far as I can see, there is no command in lme4 to save the predicted
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html#6180qlink2>
> > values (like the predict(model) function in e.g. glm).
>
>
> If you want the predictions at the observed values of the covariates you
> can use
>
>  fitted(model)
>
> > This gives the following R output: Error in predict(lmer(model)) no
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html#6180qlink3>
> > applicable method for "predict"
>
>
> > I found the same question in the R forum archives, but no answer.
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html#6180qlink4>
>
>
> > Could anybody please give me an advice how to solve this problem?
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html#6180qlink5>
>
>
> I haven't written a general method for predict applied to an lmer object
> because it is difficult to define what it should do. It is clear what
> the predictions based on the fixed effects only should be and perhaps it
> is clear what the standard errors of those predictions are (although
> that would be a case where my favorite topic of the degrees of freedom
> associated with a standard error would rear its ugly head again).
>
> It is trickier to define the predictions should be when you want to
> incorporate the random effects. If you incorporate all the "levels" of
> the random effects I think it is clear what the prediction should be.
> Defining a standard error for that prediction could be difficult - I'm
> not sure. However, I don't know what the answer should be if you only
> incorporate some of the random effects. We could define that
> unambiguously for lme models because the grouping factors were required
> to be nested. Because lmer allows for fully crossed or partially crossed
> grouping factors the concept of levels is lost. That is, there is no
> strict hierarchy in the grouping factors and we can't levels to define
> predictions.
>
> The bottom line is that I won't be able to write a predict method for
> lmer objects until I can decide what it should do, what options should
> be allowed and what the calling sequence should be.
>
> ________________________________
>
>
> R-help at stat.math.ethz.ch 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. Received on Wed Nov 29
> 18:44:37 2006
>
> *       This message: [ Message body
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6180.html#start>  ]
> *       Next message: Guenther, Cameron: "[R] Counting zeros in a
> matrix" <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6181.html>
> *       Previous message: Liaw, Andy: "Re: [R] automatic cleaning of
> workspace" <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6179.html>
> *       In reply to Fucikova, Eva: "[R] Predicted values in lmer
> modeling" <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/6158.html>
> *       Next in thread: Benilton Carvalho: "[R] predict on biglm class"
> <http://tolstoy.newcastle.edu.au/R/e2/help/07/02/10349.html>
> *       Reply: Benilton Carvalho: "[R] predict on biglm class"
> <http://tolstoy.newcastle.edu.au/R/e2/help/07/02/10349.html>
>
> *       Contemporary messages sorted: [ By Date
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/date.html#6180>  ] [ By
> Thread <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/index.html#6180>
> ] [ By Subject
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/subject.html#6180>  ] [
> By Author
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/author.html#6180>  ] [
> By messages with attachments
> <http://tolstoy.newcastle.edu.au/R/e2/help/06/11/attachment.html>  ]
>
>
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>