Dear useRs, I have a large dataset (600 to 2000 observations) that I need to cluster with a hierarchical technique, and then examine so that I can use cutree to extract various groups. I would like to visualize the results of hclust in a dendrogram, but this is obviously too much data for the normal 6 x 6 figure. Does anyone know how to make a large graph so that I can print it and view the details? I have tried changing the graph size and outputting the chart to an image of pdf file, but I keep getting an error:> jpeg(file="dendrogram.jpeg") > par(fin=c(20,20)) > plot(eq2)Error in plot.new() : figure region too large> dev.off()Please advise. Thanks, Jesse Lecy PhD Candidate in Social Science Syracuse University [[alternative HTML version deleted]]
Jesse D Lecy wrote:> > [snip] > > Does anyone know how to make a large graph so that I can print it and > view the details? I have tried changing the graph size and outputting the > chart to an image of pdf file, but I keep getting an error: > >> jpeg(file="dendrogram.jpeg") >> par(fin=c(20,20)) >> plot(eq2) > > Error in plot.new() : figure region too large > >> dev.off() > >Use the width= and height= (and possibly units=) arguments to jpeg() to specify the width and height of the graphic I would suggest that png() is more appropriate than jpeg() for this kind of graph in any case, unless you have some specific reason to use JPEG (which were designed for continuously varying images like photographs, and will produce artifacts when trying to compress a line graph). Ben Bolker -- View this message in context: http://www.nabble.com/creating-large-graphs-tf4703942.html#a13446542 Sent from the R help mailing list archive at Nabble.com.