[Wannier] Question about final spreads

[Nicola Marzari]

Thanks - I also remember that in our work on MnO we needed really
very very accurate minimizations to get to the true minimum - it's
described a bit in the 2002 PRB with Posternak (we needed 10000-50000
steps, albeit the algorithm now is much better).

So I wonder if, by continuing the minimization for many more iterations,
you would actually converge to a unique minimum.

			nicola


lan haiping wrote:
> Dear Nicola,
>  
>  Thank you.
> I do want to repeat Prof  Postermak's works on tio2.
> Last night, i also tried to change projection functions for  rutile 
> tio2. Indeed, i also found there 
> are 2 more minimum spreads founded with different stating projection 
> functions.
>  
> For rutile case, when i started with s/px/py/pz on O sites, i obtained 
> total spreads  of 14.919.  while changing projection functions to sp2/pz 
> on O sites,  i obtained total spreads of 14.4672.  
> But wannier function centers can both be classified into 4 groups.
>  
> Regards,
>  
> Hai-Ping


-- 
---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu