Re: error on using idrisi/ArcView images as masks?

["Edzer J. Pebesma" <e.pebesma@geog.uu.nl>]

Peter Schlesinger wrote:
> 
> Hello,
> I'm trying to get GSTAT 2.0f to work using IDRISI images (or even
> ArcView grids for that matter), but I keep getting errors on both of
> these. Yet the grid or Idrisi image is intact by the standards of each
> of those softwares. So, some questions:
> 1) Is the real format for IDRISI expected to be the old 6-byte or new
> 4-byte?
4-byte only.

> 2) The error I'm getting from GSTAT when fed an IDRISI image entitled
> "maskmap.img" that has a "maskmap.doc" IDRISI document file as 'maskmap'
> as designated in the manual is:
> Warning: uncomplete idrisii header in maskmap (error: '0')
> gstat: read failed on file 'maskmap'
> In the IDRISI case, my .cmd file says:
> data(ppts):'ppts.eas', x=1, y=2, v=3;
> mask:       'maskmap';
> output: 'pred.eas';

Did you check all the conditions, as mentionend in the manual
http://www.geog.uu.nl/gstat/manual/node45.html ?

> 
> If fed the same image as an ArcView gridded .flt file (and associated
> .hdr file) called a floatgrid in the online Gstat manual, I get the
> following:
> Warning: converted null-character(s) in `mask.map' to space
> gstat: read failed on file `mask.map'
> In the ArcView case, my .cmd file says:
> data(ppts):'ppts.eas', x=1, y=2, v=3;
> mask:       'mask.map';
> output: 'pred.eas';
> 

if mask.map is the arcfloat file, then there must be two files:
the file mask.map, a flat binary file with the numbers and
the file mask.map.hdr, containing the topology (ascii). No .flt file.
Does ArcView by default write the numbers to a .flt file? Can you send
me one of these .flt .hdr couples?

(the ArcGrid command gridfloat does not, as far as I know)

> What am I doing wrong?
If you get this far and want output, you should specify

predictions(ppts): 'out.pr';
variances(ppts): 'out.var';

as well, instead of the output command.

Good luck,
--
Edzer