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

---------------------------------------------------------------

After the Fall 2006 elections, Fran Miles sent me the following links
to some info about the vote ...

W-SchoolBond.html
PollingPlaceDetails_Web.xls

Darrell Massengil had an ESRI shape file of the voting precincts,
so I decided to try to put the data together with the map.
(Note that the map is not guaranteed to be 100% current with 
the data - it might be a couple years old.)

-

I used v9.2 "proc mapimport" to read in the ESRI map into a SAS
dataset which "proc gmap" can use.

I used "proc import" to read in the excel spreadsheet(s) of 
election data.

I merged the vote count data with the polling-places data,
so that I could add some text description about the polling
places when I plot the map.  I use gmap's html= option to 
display this info as html charttips/rollover-text.

I do my own "binning" of the data so I can split the bins
exactly where I want (in particular, I definitely wanted 
to have a split for above/below 50 percent).

When the 'yes' votes were less than 50%, I make those shades of red,
and above 50% (ie, the vote passing) I make those shades of green.

The bond vote seemed to have a higher acceptance rate in the 
urban/city areas (Raleigh & Cary), and found more opposition
in the rural/outlying areas.

-----

Back to Samples Index