[Ifeffit] Feff for > 175 atoms

20 Aug 2007


      Hi Eckhard,
You did the right thing. nclusx is the parameter you want to change and
you want to change it everywhere that it appears in the whole file.
Bruce's suggestion to check the results is also good. Also, just so that
you know, the code will eventually crash (or possibly not compile)  if
nclusx is set too large. There are some workarounds, but I'll leave that
for another day.
Cheers,
Josh Kas

On Mon, 20 Aug 2007, ifeffit-request@millenia.cars.aps.anl.gov wrote:
...
Send Ifeffit mailing list submissions to
  ifeffit@millenia.cars.aps.anl.gov
To subscribe or unsubscribe via the World Wide Web, visit
  http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
or, via email, send a message with subject or body 'help' to
  ifeffit-request@millenia.cars.aps.anl.gov
You can reach the person managing the list at
  ifeffit-owner@millenia.cars.aps.anl.gov
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Ifeffit digest..."
Today's Topics:
1. Feff for > 175 atoms (s440697@stud.uni-goettingen.de)
  2. Re: Feff for > 175 atoms (Bruce Ravel)
  3. Re: Feff for > 175 atoms (s440697@stud.uni-goettingen.de)
  4. Feff for > 175 atoms (appendix) (s440697@stud.uni-goettingen.de)
  5. Re: Feff for > 175 atoms (appendix) (Bruce Ravel)
----------------------------------------------------------------------
Message: 1
Date: Mon, 20 Aug 2007 10:59:25 +0200 (CEST)
From: s440697@stud.uni-goettingen.de
Subject: [Ifeffit] Feff for > 175 atoms
To: "XAFS Analysis using Ifeffit" 
Cc: Bruce Ravel 
Message-ID:
  <41184.134.76.41.165.1187600365.squirrel@webmail.stud.uni-goettingen.de>
Content-Type: text/plain;charset=iso-8859-1
Hello,
i tried to run feff with a atomcluster of 249 atoms (under linux) and got
this message:
You specified a cluster of 249 atoms for the FMS calculation.
This exceeds the hard wired limit of 175 atoms.
The cluster size was reset to 175 and the calculation will continue.
Then i made a feff-run with feff84_300.exe (under windows) with the same
249-atomcluster. This worked fine. But is there a linux-version for 300
atoms? Or is there a way to make a feff-run under linux with such big
cluster?
I would be very thankful for an answer :-)
sincerely yours,
Eckhard Bosman
---------------------------------
Eckhard Bosman
e.bosman@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
------------------------------
Message: 2
Date: Mon, 20 Aug 2007 08:15:14 -0500
From: Bruce Ravel 
Subject: Re: [Ifeffit] Feff for > 175 atoms
To: s440697@stud.uni-goettingen.de,	XAFS Analysis using Ifeffit
  
Message-ID: <200708200815.14582.bravel@anl.gov>
Content-Type: text/plain;  charset="iso-8859-1"
On Monday 20 August 2007, s440697@stud.uni-goettingen.de wrote:
...
Hello,
i tried to run feff with a atomcluster of 249 atoms (under linux) and got
this message:
You specified a cluster of 249 atoms for the FMS calculation.
This exceeds the hard wired limit of 175 atoms.
The cluster size was reset to 175 and the calculation will continue.
Then i made a feff-run with feff84_300.exe (under windows) with the same
249-atomcluster. This worked fine. But is there a linux-version for 300
atoms? Or is there a way to make a feff-run under linux with such big
cluster?
I would be very thankful for an answer :-)
Eckhard,
This is not a version problem, per se.  The issue is how the various
constants are set when Feff is compiled.  When you unpack the source
code for feff, you will see a whole slew of directories.  One of them
is called "HEADERS/".  In that directory is a file called "dim.h".
Near the top of that file, you will find a constant called "nclusx".
In the source tarball, this parameter is set to 175, but you can
change it and recompile.
Once you change that value to 300 (or whatever) and recompile Feff,
you should be able to run your problem on your linux machine.  I
suspect (but do not know for sure) that the only difference between
the windows feff84_300.exe and the normal feff84.exe is the size of
the nclusx parameter.
HTH,
B
--
Bruce Ravel  ---------------------------------------------- bravel@anl.gov
Molecular Environmental Science Group, Building 203, Room E-165
MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007
Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
Argonne IL 60439, USA                                fax: (1) 630 252 9793
My homepage:    http://cars9.uchicago.edu/~ravel
EXAFS software: http://cars9.uchicago.edu/~ravel/software/
------------------------------
Message: 3
Date: Mon, 20 Aug 2007 16:02:58 +0200 (CEST)
From: s440697@stud.uni-goettingen.de
Subject: Re: [Ifeffit] Feff for > 175 atoms
To: "Bruce Ravel" 
Cc: XAFS Analysis using Ifeffit 
Message-ID:
  <43756.134.76.41.165.1187618578.squirrel@webmail.stud.uni-goettingen.de>
Content-Type: text/plain;charset=iso-8859-1
Hello Bruce,
thanks for your fast answer. Excuse me, for the stupid question. But i
can't find the packed (linux)source-code for feff with slew of
directories. I only found one single file, the feff84.f or the
feff84.tar.gz or the feff84.zip (which contains only the feff84.f file) on
the feff-distribution webpage.
(http://leonardo.phys.washington.edu/feff/Secure-area/Bin-Source-feff8/Source...)
Thanks for your help.
Eckhard
...
Eckhard,
This is not a version problem, per se.  The issue is how the various
constants are set when Feff is compiled.  When you unpack the source
code for feff, you will see a whole slew of directories.  One of them
is called "HEADERS/".  In that directory is a file called "dim.h".
Near the top of that file, you will find a constant called "nclusx".
In the source tarball, this parameter is set to 175, but you can
change it and recompile.
Once you change that value to 300 (or whatever) and recompile Feff,
you should be able to run your problem on your linux machine.  I
suspect (but do not know for sure) that the only difference between
the windows feff84_300.exe and the normal feff84.exe is the size of
the nclusx parameter.
HTH,
B
---------------------------------
Eckhard Bosman
e.bosman@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
------------------------------
Message: 4
Date: Mon, 20 Aug 2007 16:35:21 +0200 (CEST)
From: s440697@stud.uni-goettingen.de
Subject: [Ifeffit] Feff for > 175 atoms (appendix)
To: "Bruce Ravel" ,	"XAFS Analysis using Ifeffit"
  
Message-ID:
  <53871.134.76.41.165.1187620521.squirrel@webmail.stud.uni-goettingen.de>
Content-Type: text/plain;charset=iso-8859-1
Hello Bruce,
i opened the one big feff84.f file and searched for "nclusx=175" and
changed this to "nclusx=300" in the whole file. Then i compiled it with
f77 -o and so on. I started my feff-run and it seems to be good. Is this
proceeding wise? I don't know much about such things :-/
Thanks for your help and time.
Eckhard
...
Eckhard,
This is not a version problem, per se.  The issue is how the various
constants are set when Feff is compiled.  When you unpack the source
code for feff, you will see a whole slew of directories.  One of them
is called "HEADERS/".  In that directory is a file called "dim.h".
Near the top of that file, you will find a constant called "nclusx".
In the source tarball, this parameter is set to 175, but you can
change it and recompile.
Once you change that value to 300 (or whatever) and recompile Feff,
you should be able to run your problem on your linux machine.  I
suspect (but do not know for sure) that the only difference between
the windows feff84_300.exe and the normal feff84.exe is the size of
the nclusx parameter.
HTH,
B
---------------------------------
Eckhard Bosman
e.bosman@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
------------------------------
Message: 5
Date: Mon, 20 Aug 2007 09:37:51 -0500
From: Bruce Ravel 
Subject: Re: [Ifeffit] Feff for > 175 atoms (appendix)
To: s440697@stud.uni-goettingen.de
Cc: XAFS Analysis using Ifeffit 
Message-ID: <200708200937.51359.bravel@anl.gov>
Content-Type: text/plain;  charset="iso-8859-1"
On Monday 20 August 2007, s440697@stud.uni-goettingen.de wrote:
...
i opened the one big feff84.f file and searched for "nclusx=175" and
changed this to "nclusx=300" in the whole file. Then i compiled it with
f77 -o and so on. I started my feff-run and it seems to be good. Is this
proceeding wise? I don't know much about such things :-/
That seems right to me.  As a test, you could compare the result on
linux with nclusx=300 to what you ran on windows.  If they are the
same then the two are equally right (or equally wrong!).
B
--
Bruce Ravel  ---------------------------------------------- bravel@anl.gov
Molecular Environmental Science Group, Building 203, Room E-165
MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007
Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
Argonne IL 60439, USA                                fax: (1) 630 252 9793
My homepage:    http://cars9.uchicago.edu/~ravel
EXAFS software: http://cars9.uchicago.edu/~ravel/software/
------------------------------
_______________________________________________
Ifeffit mailing list
Ifeffit@millenia.cars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
End of Ifeffit Digest, Vol 54, Issue 12
***************************************

[Ifeffit] Feff for > 175 atoms

joshua jason kas