While I disagree that a longer plot will be any clearer; I can answer your technical question.

(in R) The size of the plot is scaled to the "output device" which you can specify by wrapping the plotting code in two new statements.

pdf("filename.pdf",width=24,height=8)
#heatmap()
dev.off()

That will make a 24" wide, 8" tall file that should look nice. Depending on your environment those might be specified as centimeters or inches. Another good one is

png("filename.png",width=2000,height=800)
#heatmap()
dev.off()

That will make a 2000x800 pixel bitmap image.

While it is not directly relevant to drawing longer heatmaps in R, I have been developing alternative tools for heatmap visualisation that you might find useful:

HEAPcloud

Usage: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/HEAPcloud.pdf

Source Code: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/HEAPcloud_v0.1.zip

Example: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/HEAPcloud_v0.1/HEAPcloud.html

PHYLOmap

Usage: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/PHYLOmap.pdf

Example Datasets: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/PHYLOmap_example_datasets.zip

PHYLOmap.py: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/PHYLOmap.py

collateResults.pl: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/collateResults.pl

Path2Newick.java: http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/Path2Newick.java

Best Wishes, Umer

I appreciate that this has been a long time since your question, and may not directly answer your question. But I just developed a Julia-based GUI tool to plot heatmaps based on Python's Seaborn package, just wanted to share it here. Very minimal/no coding is required for this. This is the link:

https://github.com/srgk26/easy_plotting.app