[Wien] x nn (bug?)

[Peter Blaha]

Hi,
Thank you very much for your mail. Yes indeed, I increased  "dstmax" in nn.f,
in order to get even for supercells with large vacuum proper nn results.
However, this causes the error you reported. One needs to increase also the
number of unitcells when calculating nn distances.

You can fix the problem by changing the triple-loop
          DO 180 I1=-4,4
          DO 180 I2=-4,4
          DO 180 I3=-4,4
to
          DO 180 I1=-5,5
          DO 180 I2=-5,5
          DO 180 I3=-5,5

The update will be on the web soon.

Regards

> I updated the newest wien2k code (after 04-Dec-2003), there are no error
> messages during the installation.
> I used "x nn" to test some structure files both with the old version
> (downloaded at about 11-Sep-2003) and the new version (downloaded after
> 04-Dec-2003).
>
> For the TiC case, there are no errors for both versions.
> For the fe8n1 case (space group no. 225, please see attached structure
> file), there are no errors for old version, but there are errors for the new
> version as follows:
> --------------
> ERROR: Mult not equal. PLEASE CHECK outputnn-file
> ERROR: ityp not equal. PLEASE CHECK outputnn-file
> .
> .
> --------------
> When I used the "x sgroup" of the new version to test the same fe8n1.struct,
> then no error massages, it gave me the right space group.
> Was something wrong with my complied "nn" program?  Or there are bugs in the
> new nn program.


                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------