[Wien] SECLIT error

[Laurence Marks]

I am sorry, but I think the suggestions that were made are not fully
correct. While this problem can be due to a divergent calculation,
this is certainly not the only source.

1) If they are due to a large :DIS and/or :PLA and the calculation
diverging I strongly suggest that you update to the latest version
since this problem has been largely solved. If for some reason you
have to use an old version, check :PLA and the angle and search the
mail list for information about how to adjust the scaling of the plane
waves relative to the CLM's.

2) If this is not the issue, look to see if the iteration is giving
QTL-B warnings. You may need to use -in1new if you were not before.
Note: I recommend that you never use -in1new straight from dstart as
this can lead to instabilities. (I think the latest version has this
trapped.)

3) Consider changing the number of k-points slightly or adjusting the
linearization energies; please read the documentation carefully.

4) Check in case.output1_X for the lines which start with "LO
COEFFICIENT:". If you have two terms for a given L and C=0 for both
the lo and normal LAPW term have become linearly dependent. This has
been fixed in the latest version; with earlier versions you need to
change the linearization energies by hand.

On 9/6/07, jadhikari at clarku.edu <jadhikari at clarku.edu> wrote:
> Dear S. Jalali,
>
> Thank you very much for the suggestions. I would consider changing mixing
> parameter first.
>
> Have a nice time.
>
> Subin
>
>
> > This may be taken as an indication to a diverged calculation. You would
> > first make sure whether or not there is a big jump in your calculated
> > charge distance  between two immediate iterations, i.e. :DIS:
> > grep :DIS *scf
> >
> > If you detect a large fluctuation in the :DIS after some iterations, you
> > may correspond it to your selected mixing parameters and/or your performed
> > linearization energies.
> > Maybe in this case you can solve the problem by reducing the mixing
> > parameter in the case.inm file and/or increasing the number of iterations
> > N in -in1new N (if you use this flag at all).
> >
> > Your,
> > S. Jalali.
> >
> > jadhikari at clarku.edu wrote: Dear WIEN users,,
> >
> > In the 27th cycle, LAPW1 step, the calculation crashed with this error-
> >
> > SECLIT- Error in Cholesky
> >
> > I would be grateful for any suggestion for treating this error.
> >
> > Thanking you.
> > Subin
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
> >
> >
> > ---------------------------------
> > Luggage? GPS? Comic books?
> > Check out fitting  gifts for grads at Yahoo!
> > Search._______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>


-- 
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60208, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
Web: www.numis.northwestern.edu
EMM2007 http://ns.crys.ras.ru/EMMM07/
Commission on Electron Diffraction of IUCR
www.numis.northwestern.edu/IUCR_CED