[Wien] Still problem with CLMSUM

L. D. Marks L-marks at northwestern.edu
Sun Mar 12 14:24:55 CET 2006


The values in the case.clmval should not be zero. What you sent suggests 
that the problem is not in mixer, but somewhere earlier. The case.clmval 
files are produced by lapw2 and some related codes, so the next issue 
is to determine whether the problem is in lapw2, the related codes or 
earlier in lapw1.

a) Look at all the clmval files, i.e. "grep -e PW *.clmval_* -A 5". Do 
these all show zero's for the PW's ? If they don't, this implies a problem 
in sumpara which adds them.

b) Look at case.output2, do the bandranges look reasonable?

c) Do a "grep -e :CHA *.output2_*", is lapw2 giving you charge?

d) Look at your case.scf1_X (X=1,2... for as many processors as you used). 
Do the eigenvalues look reasonable.

e) You mentioned something about it working with a smaller GMAX. Can you 
run it with a smaller GMAX, and then increase the value to find when/if it 
breaks. (Save the calculations before increasing GMAX.) This is quite safe 
to do.

I would suspect that you have a problem related to the very large number 
of your plane waves, but have no idea which code and until you narrowed 
down more it's hard to help.

-----------------------------------------------
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60201, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
http://www.numis.northwestern.edu
-----------------------------------------------



More information about the Wien mailing list