[Geodata] [Tiger] A few interesting observations on the Tiger2007fedata

[John P. Linderman]

I'm (finally) getting back to the question posted earlier

Stephen Woodbridge <woodbri at swoodbridge.com> asked (with some editing)
on Mon, 30 Jun 2008 17:08:21 -0500 (18:08 EDT):

> One problem on interpreting the data for geocoding will be how to decide 
> what the various address ranges mean. ie: so they overlap each other are 
> the stacked like |-- range 1 --+--- range 2 --|, or are they just random 
> house numbers that sit on the street given some other range along the 
> street, etc.

Since the shapefile distribution has no explicit ordering imposed
on multiple address ranges for the same side of the same edge,
we have no way of deciding whether range 1 comes before or after
range 2.  If we merge overlapping odd/even ranges, we at least
eliminate the ambiguity.  So, on that basis alone, I'd favor
merging wherever possible.  -- jpl