running i.in.erdas

Jim Stafford jimstaff at freenet.columbus.oh.us
Sat Feb 11 11:36:13 EST 1995


On Fri, 10 Feb 1995, Reuben Wee wrote:

> Hi, I wrote a couple of days ago regarding converting erdas file to grass 
> file format. I'm running Grass 4.1 update 3 on Linux 1.1.59. I 
> encountered some problems here when running i.in.erdas. Here is a log of 
> what happened...
> 
> -----------
> 
> tribe # i.in.erdas input=hingsat.lan output=dataset1
> HEAD74
> pack type------ 0
> number bands----------- 1792
> number cols,rows------- -1240662016, -1240662016
> starting coordinate --- 16777216, 16777216
> map type -------------- 0
> number classes -------- 1
> area ------------------ 0.000000 N
> map coordinate -------- 0.000000, 0.000000
> pixel size ------------ 0.000000, 0.000000
> 
> Error in memory allocation
> 
> -------
> 
> The person who passed me this file ensured me that this was a 7 band 
> file... As, I'm totally lost with Grass and GIS in particular, I wouldn't 
> know how to make head or tails out of the raw file (hingsat.lan - and the 
> 90 Megabyte size makes my bone chill as to try and see where the fault 
> lies). I noticed that in the console output, i.in.erdas interprets the 
> file as having 1792 bands (which is as far as I know impossible). And it 
> crashes with a memory allocation error... Before I conclude that I have a 
> faulty data file, is there any other thing I should be aware of ?
> 
> Roobs
> 
> UNIMAS - Universiti Malaysia Sarawak
> ====================================================== ora et labora =
> Reuben Wee (office: roobs at fit.unimas.my
> 	    home  : reuben at tribe.pc.my
> 	    others: roobs at nabalu.flas.ufl.edu
> 	    URL	  : http://www.fit.unimas.my:8080/users/roobs
> ================================================================i=l=m=
> 
Dear Reuben,
I also have had problems with erdas import routine (r.in.erdas,
specifically) some of which were alleviated with the following from Janne
Soimasuo.

"Check Gmakefile in .../src.contrib/SCS/raster/r.in.erdas/cmd

There is a line:
#EXTRA_CFLAGS = -DATT_386

#-mark should be used (commented out) with non intel based systems (like
suns), but with e.g. linux you should define ATT_386-flag.
This is true for both r.in.erdas and i.in.erdas."

Jim Stafford
email:jimstaff at freenet.columbus.oh.us





More information about the grass-user mailing list