[Wien] how to extact...Parabolfit needs small change

[Peter Blaha]

Yes, I agree the current parabolfit programs has a couple of problems:

a) For the 4D-case, it would work only with variations of angle alpha, 
not beta or gamma. I've therefore added an option -a/b/g which lets you 
select which angle you want to consider in the 4D fit.

b) Yes, the scf selection was not consistent. Without the label:
  -scf '*xxx*scf'   (where xxx stands for a unique identifyer) it was 
working properly with the default files, but not otherwise.

The atached parabolfit_lapw -h:

parabolfit_lapw [ -t 2/3/4 ]  [ -f FILEHEAD ] [ -scf '*xxx*.scf' ] [-a/b/g]

should fix both problems.

PS: Nevertheless: for a 4D case and only 15 cases, it is just the 
absolute minimum to make a harmoncic fit. Every single "noisy" energy 
will completely change the results and I'd trust those results at best 
to find an approximate minimum.


On 11/18/2017 01:37 PM, Lyudmila wrote:
> 18.11.2017 15:21, Lyudmila wrote:
>> Better try and find the place in parabolfit that determines which part
>> of the output should be rewritten in the ZrO2.ene and ZrO2.latparam. I
>> think that this line should be changed somehow:
>>   grepline :lat '*mon*.scf' 1 -s| cut -c27->$file.latparam
> I have looked through this: the file latparam contains all 3 angles, and
> eosfit6 in a 4-d case takes only one angle (the first one, maybe, that
> is, the forth column).
> So two variants:
> 1) to move the column of changing angles to a proper place.
> That is, you are doing parabolfit, than change columns in the file
> latparam, and do 'x eosfit6'
> 2) you are doing parabolfit, than do 'x eosfit6' and answer that you
> have 6 dimensions - and it will consider 6 columns, in the output you'll
> see much more variables.
>
> By the way, my new version of parabolfit has a misfit: version 3 (abc
> change) looks like
>       grepline :ene "*abc*$scf" 1 -s| cut -c40- >$file.ene
>       grepline :lat "*abc*$scf" 1 -s| cut -c27->$file.latparam
> And program does not have $scf here. I think it should be changed to:
>       grepline :ene "*abc*scf" 1 -s| cut -c40- >$file.ene
>       grepline :lat "*abc*scf" 1 -s| cut -c27->$file.latparam
>
> Best wishes,
> Lyudmila Dobysheva
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/TC_Blaha
--------------------------------------------------------------------------