One way of solving you problem is to assign each base class with a 'Magic'
number. You could use factors of 2. e.g. 1,2,4,8,16,32,...,1024,2048... etc.
The neat thing with these numbers is that if if you add them they act as a
logic OR. And from the result you van always deduce which were the origional
base class numbers. Exactly like the bits of binary number.
An obvious limitation however is the number of classes that you want to
assign. 32 bit numbers are standard so 32 classes are no problem...
Roeland
----- Original Message -----
From: Robert Crossley <[EMAIL PROTECTED]>
To: MapInfo List <[EMAIL PROTECTED]>
Sent: Monday, April 03, 2000 2:35 AM
Subject: MI: Vertical Mapper Polygon overlay
> I am trying to intersect a slope category map with a vegetation fire risk
> map to produce a third coverage showing unique combinations of slope risk
> and vegetation hazard. The resultant risk is not a simple mathematical
> function of the two, and we wish to assign classes and descriptions of
risk
> manually for each combination.
>
> The tables contain very complex polygons and the approach of splitting one
> coverage by the other, or using a program which I have written that does
> this in steps will not finish in this lifetime.
>
> I have been trying to use vertical mapper to do much of the analysis to
> date, but am now stuck.
>
> Vertical mapper does not allow grid calculations using character based
> fields, so I cannot simply create new classes based on the two character
> based grids.
>
> If I try to use numbers instead of characters to designate classes and do
> calculations to give unique results eg. use prime numbers for the base
> classes and multiply the grids, or multiply one coverage by 10 and add the
> second; I get interpolated numbers for each layer along the edge of each
> class area that no longer directly point to any one of the previous
classes
> (thus I cannot guess what combination of base layers that this area came
> from). This approach would probably work if there were rounding functions
> or integer functions available in the grid calculator.
>
> I am now selecting each combination and selecting out areas that have each
> unique combination of each of the two grids, eg. make areas where vege
class
> of 1 and slope class of 1 = 1 and other areas be null. I will have to do
> this 25 times (once for each combination of classes, and then splice the
> partial results together. This is a pain to do, and its success will be
> dependant on how null values are treated when splicing grids.
>
> Any suggestions?
>
> R
>
> Robert Crossley
> Trinity Software
> 10 Trinity Street
> CAIRNS 4870 AUSTRALIA
> Ph: 61-7-40314877
> [EMAIL PROTECTED]
>
> ----------------------------------------------------------------------
> To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
> "unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
>
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
