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

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

This is my SAS/Graph imitation of the following plot:

http://www.livebaltimore.com/nb/hs/census/
http://www.livebaltimore.com/UploadedImages/ImageGallery/census_graph.gif

Note that this graph is a bit misleading, because the y-axis doesn't start at zero!
Therefore this graph "exaggerates" the drop in population -- it you just glance
at the graph, it gives you the impression that the population has dropped by
about 50% since the 1950s.  You would think that the city of Baltimore would
not want to exaggerate the population decrease in that direction(?)

Note also that the x-axis (year) tickmarks aren't evenly-spaced ...
First it's by 10-years, then by 1-year, and then by 5-year.

The person reading this plot will have to really be on-the-ball to 
make sure they know what they're looking at!

-----

Oh well ... anyway, here's the SAS/Graph "tricks" that were used ...
I added a variable called 'range' to the data, and then I used gchart's
subgroup=range to color the bars as desired in each range.
I annotate the text on the bars (using angle=90 and position='4' to get
the text positioned just right).  I also annotate a white box onto the
graph to cover up some of the reflines, and make room for the titles
which are also annotated on the graph.


Back to Samples Index