Re: MI area calculations broken

[Warren Vick, Europa Technologies Ltd.]

Hello Mark,

> I have been making comparisons of the area of various objects, and have
> found MapInfo's area calculating function to be seriously broken, at least
> in some projections (Polar Azimuthal projections).

MapInfo Pro's area calculation is independent of the current map projection. It just 
assumes a 
spherical Earth, with the option of Cartesian in v5.5. You can see this by the fact 
that the MB 
area() function does not have any coordsys or map window ID clause.

The problem you are having can be seen by simply reverting to a rectangular 
projection. One of 
your triangles crosses the 180 degree meridian. While this looks perfectly fine on the 
polar 
projection, it is in fact implemented by a massive triangle crossing the entire world 
and this is the 
reason for your outsized area. Try breaking your trans-meridian region into a two 
polygon region, 
broken at 180deg. You should find that the area are then equal, but they may still be 
incorrect. I 
rarely trust MI's distance and area calculations these days.

One additional point. Having polygons with such long edge lengths doesn't really make 
sense 
since they are straight lines no matter what the projection is. It's better to decide 
on an 
acceptable resolution for your "lines" and insert nodes accordingly. You can then 
decide whether 
these lines are great circles, rhumb lines, etc. and be in better control of the areas 
definition.

Regards,
Warren Vick
Europa Technologies Ltd, U.K.
http://www.europa-tech.com
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]