The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <masayoshi.mikami@gmail.com>]

Dear Rick,

The simplest (but least beautiful) way might be "up to ABINIT"; when one run ABINIT with wrong input (e.g. wrong natom), ABINIT will complain in the log file and stops. Then, he/she will notice which parameters should be corrected... ("natom should be xx" etc.)  Sometimes, I corrected my input files by this way. ;-) 

The most beautiful way would be, as you said, computing relevant parameters (i.e., "natom", "typat", "ntypat") from a CIF file (space group, inequivalent atoms...)  The "natrd" should be the number of inequivalent atoms. In principle, "typat" and "natom" can be determined from the multiplicity of each Wyckoff position.
In most cases, we notice such information before running ABINIT, by drawing/visualizing crystal structure with appropriate software...

In passing, the following page might be interesting for crystallographical knowledge: 

- http://www.cryst.ehu.es/ (e.g. "WYCKPOS")

- http://www.ehu.es/ocs/index.php/fmc/CoL2009/schedConf/presentations (Presentations in "Crystallography Online: International School on the Use and Applications of the Bilbao Crystallographic Server")

Good continuation,
Masayoshi

2009/12/5 Rick Muller <rpmuller@gmail.com>

Thanks, I am even further in your debt. This indeed does work.

Is there a way to know from the CIF file alone that one has to change natrd to 2? I might have figured this out for spacegroup 227, but if I'm using an arbitrary spacegroup, how do I know? Do I just need to keep a big database of every spacegroup, and the corresponding values? I would like to be able to set this up to use a minimum of user intervention.

I'll spend some time reading through the relevant test files, as you suggested.

Thanks very much for your help with this. 

On Thu, Dec 3, 2009 at 7:19 PM, Masayoshi Mikami <masayoshi.mikami@gmail.com> wrote:

Dear Rick,

There are several errors in the input. I have corrected it, as shown below.
(Other parameters may be tuned... but they are not the problems here.) You
might want to check test cases using "spgroup" "natrd" in ~ABINIT/tests/*.

Good continuation,
Masayoshi

# CIF block 9011998 converted by cif2abin.py
ixc         11
ecut        8.0
toldfe      1.0d-6
diemac      12.0

acell       5.430941 5.430941 5.430941 angstrom # default (a.u.), not angstrom

angdeg      90 90 90
spgroup     227
brvltt      -1

kptopt      1
ngkpt       3 3 3

ntypat      1
znucl       14

#natom   1  (2 atoms in the primitive unit cell )
natom    2
typat       1
natrd       1  # to be added

xred
0.00000 0.00000 0.00000

--
Rick Muller
rpmuller@gmail.com
505-750-7557