This whole post originates from this stackoverflow question. The original poster(OP) was having a hard time putting multiple `plots` into the same canvas. It looks like he was using the base plot function and I assumed the par(mfrow) will be enough in that case. However, it turned out this is a way more interesting question than I first realized and here are all the `aha`s.
He was using a library called DoseFinding , which is
a package provides functions for the design and analysis of does-finding experiments(for example, pharmaceutical Phase II clinical trials).
In there you can use the base plot function to plot the return object from the function `fitMod` (the does response model). You can see the return class of fitMod is called “DRMod” (drug response model) and the plot class is called “Trellis”.
Actually, Trellis is a whole visualization framework developed by the Bell Lab and AT&T Research. Here is another post to understand post to give you a basic understanding of lattice and trellis.
I have working working with the base plot functions like boxplot, hist, plot.. and also ggplot2 from Hadley Wickham for a while, and I am surprised to find out how easy it is to use lattice package to draw plots taking multiple variables into consideration.
There are also so many different plot methods in lattice, which you can spend more time exploring.

