Figure margins too large on TraMineR plot

Question

Im trying to graph types as shown here http://traminer.unige.ch/preview-typology.shtml

I can only fit 8 types in my screen until i get the error Error in plot.new() : figure margins too large. That´s as far as my UI goes, I can't make the graph interface any taller.

I'm trying to make more types, is there any way I can do this?

This is the plot I'm trying to do.

seqIplot(f3.seq, group = cl1.4fac, sortv = "from.end",with.legend = "none", xtlab=c(rep(18:29, each = 1)))

Answer 1

The 'Figure margins too wide' error is related to the graphic device you are using. It means that the margins do not leave enough place for the figure itself.

There are several possible workarounds (that can possibly be combined).

1. Changing the number of columns of the graphical area by means of the cols argument of the seqplot function.

2. Reducing the size of the tick marks labels (cex.axis argument), of the title (cex.main argument) and suppressing axis labels.

3. Suppressing the y and x axes (axes and yaxis arguments).

4. Changing the values of the margins (par(mar=...)).

5. Changing default parameters of the graphic device, e.g. play with the width and height values of pdf() or png().

I illustrate the first four solutions below using the mvad data that ships with TraMineR.

library(TraMineR)
data(mvad)
mvad.labels <- c("employment", "further education", "higher education", 
                 "joblessness", "school", "training")
mvad.scode <- c("EM", "FE", "HE", "JL", "SC", "TR")
mvad.seq <- seqdef(mvad[, 17:86], states = mvad.scode,
                   labels = mvad.labels, xtstep = 6)

diss <- seqdist(mvad.seq, method="LCS")
mvad.clus <- hclust(as.dist(diss), method="ward.D")
k <- 9
mvad.cl <- cutree(mvad.clus,k=k)

## changing text size and number of columns of graphical area   
seqdplot(mvad.seq, group=mvad.cl, border=NA, cols=3, 
         cex.axis=.5, cex.main=0.6, ylab='', xlab='')

## default margins are: par(mar=c(5, 4, 4, 2) + 0.1)
## setting 0 margins except for top
opar <- par(mar=c(0, 0, 2, 0) + 0.1) 
seqdplot(mvad.seq, group=mvad.cl, border=NA,  
         axes=FALSE, yaxis=FALSE)
par(opar) ## resetting default mar