Click here to see the SAS code.
Click here to see the example.


To get the "data" for this map, I had to look at the original map,
and use their pre-binned cagetories (legend values).  I would have
preferred to have the actual data for each county, and then I would
have done quintile-binning such that 1/5 of the counties were in 
each color/category ... as it is, the legend bins look to be even
divisions of "10", which is mathematically convenient, but don't
necessarily lend themselves to the data.

Also, in the original map, it is hard to distinguish the colors
(especially since the 2 counties with the darkest red are very
geographically small counties, which makes their color look 'less').
Therefore, I did my own version of the map and used colors that are 
easier to distinguish (from - but also did one in
the original version's color scheme to show SAS could do it.

I chose to make my map smaller than the original, so it would better
fit on 1 screen ... that would make the county labels a bit small/cluttered,
so I left off the labels and used html rollover-text for the county names.

I used "proc gproject" to "clip" the small/inset rectangular map area.
I used "proc greplay" to put both maps on the same page.

Back to Samples Index