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

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

See my blog for more information!


This is a SAS/Graph proof-of-concept, showing how Proc Gmap 
could be used to display social-media data, geographically...

-----

I have seen copies of the original map, in varying sizes, all over 
the Internet.  I tracked down the following article on the 
Facebook site, which I assume is the "origin" of the map:

https://www.facebook.com/notes/facebook-data-science/nfl-fans-on-facebook/10151298370823859

They say "At Facebook we have about 35 million account holders in the 
United States who have Liked a page for one of the 32 NFL teams".
"By considering the physical locations of NFL fans, we can construct 
a map of the top team for each county in the US"

-----

I could not find the raw data anywhere, therefore I estimated
the data for each county based on the apparent color in the 
Facebook map.  This was a bit challenging, even with a screen
color-reader (Pixeur), because the colors in the legend and 
map are not consistent, and the color even varies with several
shades within each county.  Combine that with the face that 
several teams are represented with nearly the same color 
(such as dark red, purple/blue, or gray).  Therefore some 
counties are my "best guess" and might not be correct.

Improvements:

The main improvement in my SAS/Graph map is that SAS uses the
exact same color in the legend and the map areas (no shading/smudging/
or blending).  Also, I added html mouse-over text, so you can hover
your mouse over a county and see the county name & favorite team
in plain text - this makes it *much* easier to determine the 
teams, rather than trying to visually find the matching color in 
the legend.
 
Also, I put the date in a footnote on the map - this way, years from 
not people will know that this was a snapshot of data from the past,
rather than the current data from their year.

Back to Samples Index