[Ifeffit] nclusx=500

[joshua jason kas]

Hi Eckhard,
For now I will only answer your question about nclusx since I have to 
investigate the neg dens situation.
You are running out of memory when you set nclusx=500.
There are two possibilities here.
1. You don't have enough RAM.
 	I think the problem is that feff84.f on the web has some funny
 	defaults that need to be changed. nex is another parameter which
 	gives the number of energy points. I am guessing that it is set to
 	450 in your version. Set it to 150. This was the default for
 	feff82, and should be the default for feff84. I will change the
 	one on the web as well.

2. Your stacksize is not set to unlimited.
 	To figure out what the stacksize is set to, type
 		ulimit -s
 	If it doesn't return unlimited, type
 		ulimit -s unlimited

 	The above commands are for bash. I think tcsh might use a
 	different command (limit maybe?).

For the negative density stuff, could you send your feff.inp file?
Cheers,
Josh Kas

>
> Hello,
>
> i have 2 questions.
>
> I read in the log1.dat file (from a xanes-calculation, see also attachment)
> negative dens  10  2
> negative dens  20  2
> negative dens  30  2
> negative dens  40  2
> What does this mean? negative dens=negative density ?? Is this a message i
> have to be worried about?
>
> The second problem is a program break. I modified the entry nclusx=175 to
> nclusx=500 in the feff84.f to make a xanes-calculation with a 328-Atoms
> cluster (or maybe later more). The compilation (f77 -o feff84_500
> feff84_500.f) finished without messages and i started the calculation with
> feff84_500 but then the prompt answered with:
> $ feff84_500
> Killed
> $
>
> I did the same procedure with nclusx=400 and compiled it. With feff84_400
> the calculation works... Does anyone know if there is a problem with
> nclusx=500?
>
> Thanks for your help and time.
>
> sincerely yours,
> Eckhard Bosman
>
> ---------------------------------
> Eckhard Bosman
> e.bosman at stud.uni-goettingen.de
> +49 (0)551-39-14441
> Raum: E0.104
> Institut f?r R?ntgenphysik
> Friedrich-Hund-Platz 1
> 37077 G?ttingen
> Germany
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: log1.dat
> Type: application/octet-stream
> Size: 8948 bytes
> Desc: not available
> Url : http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20070821/cf55033c/attachment-0001.obj
>
> ------------------------------
>
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>
>
> End of Ifeffit Digest, Vol 54, Issue 13
> ***************************************
>