Not triclinic! [was: Re: [Wien] Problem creating struct files from VASP contcar files]

[Gus Hart]

At 12:02 PM 1/13/2004 +0900, you wrote:
>On Tuesday 13 January 2004 06:56, Gus Hart wrote:
> > Doug du Boulay and Peter Blaha have been thinking in WIEN2K mode to long...
>
>Granted, but it is not just Wien2k. For better or worse, I would say almost
>all crystallographic software works purely with standard cell settings,

All first principles codes that I have used (about 6) EXCEPT this LAPW code 
can accept primitive unit cells, even if they have other input options.

>
>because they're standards.

I have to take issue with that...

Many of us in alloys--not just myself--deal with predicted structures and 
hypothetical structures predicted by model Hamiltonians. These of course do 
not come out in "standard" non-primitive settings. VASP and even WIEN2K 
(under the hood) prefer to work with the primitive cell for efficiency 
reasons. Many of us would prefer codes that don't assume so much and let us 
just enter the primitive cells.

To imply that there isn't a market for programs that are smart enough to 
deal with primitive cells seem a bit of an over-generalization to me. 
Besides, dealing with symmetries for arbitrary primitive cells isn't hard 
(most of us wrote codes to do that as grad students...) What's hard is 
taking a perfectly respectable primitive unit cell and try to guess how to 
turn it into a "standard" one for WIEN2K.

I appreciate Flourent's last response. Hopefully that will get us some where...

All the best,
Gus H.

>  If you choose to use a setting of lower metric
>symmetry, thats the price you pay.
>
> > If one has a primitive cell (say from VASP), how do you convert it to a
> > struct file? Many cases I have done for base-centered monoclinic are very
> > difficult. One took me several days before I finally found the *magical*
> > transformation...
>
>Not a specific answer, but maybe you can find a pointer to an algorithm here:
>http://xtal.sourceforge.net/man/creduc-desc.html. For an indirect method, if
>you could produce a CIF then PLATON http://www.cryst.chem.uu.nl/platon/
>could probably identify the correct cell choice and convert all sites for you.
>
>Good luck with your transformations.
>Doug
>
>_______________________________________________
>Wien mailing list
>Wien at zeus.theochem.tuwien.ac.at
>http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

++++++++++++++++++++++++++++
Gus Hart   (Rm. 312)
Department of Physics and Astronomy
PO Box 6010
Flagstaff AZ 86011-6010
tel. (928)523-0426
fax (928)523-1371
gus.hart at nau.edu
http://www.physics.nau.edu/~hart