[R] multiple hexbin plots with varying greatest densities

[Thomas Lumley]

On Mon, Mar 26, 2012 at 4:59 AM, Wayne Gray <wgray.999 at gmail.com> wrote:
> Greetings:
>
> I have multiple hexbin plots with varying greatest densities.
>
> Right now, the data on each plot varies from 1-256 levels of density. The problem with that is that in Plot A the data are more scattered across a 2 x 2 grid whereas in Plot B the data are more concentrated in fewer cells. Hence, Plot A has a few very dense grid cells and plot B has nothing as dense as plot A.
>
> (In Plot A the legend shows that the range of densities per cell is 1-700 whereas for Plot B it is 1-453. Both plots use the full range of available colors with 256 gradients.)
>
> I want people to be able to compare the two plots and being able to see the differences in variability of the densities is one of the things I want them to see. AT PRESENT, each plot uses the full range of densities; hence, this makes it seem as if each plot has the maximum highest and minimum lowest densities.
>
> How can I do this?

With the 'lattice' or 'centroids' styles you can use the maxarea=
argument to plot.hexbin or to grid.hexagons.

survey::svycoplot() is an example of this: you can either have the
same scale across panels or have each panel use the full range of
sizes.

   -thomas

-- 
Thomas Lumley
Professor of Biostatistics
University of Auckland