r.infer

[/S=L.LEVIEN/OU=R05K/PRMD=USDA.FS.X400/ADMD=TELEMAIL/C=US/ at sprint.com]

I had a very strange problem using r.infer.  I am using slope and 
aspect maps created from DEM's and brought into GRASS.  Upon running 
r.support in prepration for running r.infer I specified the cols and 
rows and UTM's which correspond to the area. r.infer worked on only 
1020 rows even though I had 2176 and specified it as such.  I 
eventually changed the WIND, DEFAULT_WIND, and cellhd files of each 
to eliminate norhting and eastings (since after I ran r.infer many 
times differently, the rows and cols were automatically changed as 
were the UTM's), and put in rows for northing and cols for easting, 
with a resolution of 1, 1.  This worked...does any one know why?  I 
am just curious so that I can try to follow this logic in the future!
Thanks!
lisa at frrap1.cdf.ca.gov