I'm sure this has been addressed in detail in the
past, but I've never confronted this problem
before and I can't find a reasonably simple
solution using MB.

The situation is: I have some points that were
geocoded to zip centroids because they wouldn't
geocode to addresses. In some cases there are over
100 objects at the same zip code centroid
coordinate.  

The objective is:  I want to disperse them around
the zip code polygon in which they occur so
they'll approximate the dispersion of the
accurately (to street address) geocoded points and
stay within the polygon.  (The cooler objective
would be: I would like to get fancy and disperse
them "proportionately," that is by some
calculation of the relative densities of points
throughout the zip code area, but right now my
objective is more limited.)

The problem is:  I have written an MB utility that
randomly disperses the points within the MBR.  The
problem with the MBR is lot of irregular polygons
do not have much to do with the MBR except that it
describes a rectangle with max and min x and y.  I
get results where about half of the points
dispersed fall within the polygon; the other half
are elsewhere in the MBR. 
I'd like the points to fall in the polygon, not in
its MBR. BTW, I've used the point dispersion
utility, but don't like the results very much. Are
there solutions to getting points in polygons by
identifying where the polygon is and plunking the
points down in and not out?

Thanks,

Tim Rood
Ravenna Planning Associates
Seattle, WA
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]

Reply via email to