[R-sig-ME] Odd plot.lme() behavior

[ONKELINX, Thierry]

Dear David,

This is probably due to the fact that either YES or NO is NA in your
dataset and you included the na.action = na.omit argument in your model.
Without that argument in the model I get the plot with points.
My guess is that lme() omits rows only if they have at least one NA
value in a variable that is present in the model (which neither YES and
NO are). But that the plot function looks for missing values in the
entire row. Hence YES and NO are included and thus all rows omited.

HTH,

Thierry


------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature
and Forest
Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
methodology and quality assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium 
tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be 
www.inbo.be 

To 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 Tukey

-----Oorspronkelijk bericht-----
Van: r-sig-mixed-models-bounces at r-project.org
[mailto:r-sig-mixed-models-bounces at r-project.org] Namens David Daniel
Verzonden: dinsdag 20 januari 2009 6:47
Aan: r-sig-mixed-models at r-project.org
Onderwerp: [R-sig-ME] Odd plot.lme() behavior

Hi all,

I've come across what appears to me to be strange behavior in the  
plot.lme() method.  In the example code below, the first plot() call  
gives a plot with no points plotted.  But the second plot() call gives  
a plot as expected (with points).  The only difference is that in the  
second case, lme() uses a data set not containing the variable, YES,  
but this variable is not referenced in any of the functions used in  
the code.

Even more odd is that if a seed of 1 is used in the call to  
set.seed(), the second call to plot() gives an error message not given  
when the seed is 2 (the seed only affects which values of variables  
YES and NO are NA's, and these variables are not used):

|> plot(lmeB,  Group ~ resid(.,  type="p"))
|  Error in `[<-.data.frame`(`*tmp*`, , ".y", value = c(1, 1, 1, 1, 1,  
1,  :
|  replacement has 36 rows, data has 80

But if *both* variables YES and NO are removed from the data set used,  
the plot works fine with a seed of 1.

Have others seen this?  Is this expected behavior?

I'm on a MacBook Pro 2.16 GHz Core 2 Duo, with OS 10.5.5, R version  
2.7.0, package nlme version 3.1-89, and package lattice version  
0.17-8.  The sessionInfo() output is below.

Thanks,
-David

#---------------------------------------
library(lattice);
library(nlme);

# Save the user's random seed
save.seed <- .Random.seed


#---------------------------------------
set.seed(2, kind = NULL)

n <- 10	# Number of subjects in each Group
k <- 4	# Number of phases

Group <- factor(c(rep("A",n*k), rep("B",n*k)))
m <- length(Group)/k		# total subjects

Gender <- factor( rep(c(rep("M", k), rep("F", k)), n) )
ID <- rep(1:m, rep(k, m))
phase <- rep(1:k, m)
YES <- rep(ceiling(12*runif(m)), rep(k, m))
NO <- YES
YES[YES > 6] <- NA
NO[NO <= 6] <- NA
PB <- rnorm(m*k) + phase + as.numeric(Group)
plotData <- data.frame(cbind(ID, PB, Group, Gender, phase, YES, NO))

#---------------------------------------
# Removing either YES or NO from the data.frame gives the correct plot
#---------------------------------------
plotData2 <- subset(plotData, select=c(ID, PB, Group, phase, YES, NO))
lmeA <- lme(PB ~ Group + phase,
					random= list(ID= pdDiag(form=
~phase)),
					data=plotData2,
					na.action= na.omit)
plot(lmeA,  Group ~ resid(.,  type="p"))

#---------------------------------------
# Removing either YES or NO from the data.frame gives the correct plot
#---------------------------------------
plotData3 <- subset(plotData, select=c(ID, PB, Group, Gender, phase,  
YES))
lmeB <- lme(PB ~ Group + phase,
					random= list(ID= pdDiag(form=
~phase)),
					data=plotData3,
					na.action= na.omit)
plot(lmeB,  Group ~ resid(.,  type="p"))

#---------------------------------------
# Restore the user's random seed
.Random.seed <- save.seed
#---------------------------------------



 > sessionInfo()
R version 2.7.0 (2008-04-22)
i386-apple-darwin8.10.1

locale:
en_US.UTF-8/en_US.UTF-8/C/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

loaded via a namespace (and not attached):
[1] Matrix_0.999375-9 grid_2.7.0        lattice_0.17-8     
lme4_0.999375-20
[5] nlme_3.1-89

----------------------------------
David Daniel
Associate Professor
University Statistics Center
New Mexico State University

ddaniel at nmsu.edu

_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

Dit bericht en eventuele bijlagen geven enkel de visie van de schrijver weer 
en binden het INBO onder geen enkel beding, zolang dit bericht niet bevestigd is
door een geldig ondertekend document. The views expressed in  this message 
and any annex are purely those of the writer and may not be regarded as stating 
an official position of INBO, as long as the message is not confirmed by a duly 
signed document.