[gpaw-users] MethfesselPaxton, missing setups

[Marcin Dulak]

Hi,

please see how to generate and use the setups: 
https://listserv.fysik.dtu.dk/pipermail/gpaw-users/2010-December/000583.html
The parameters for missing elements are:
 'Y':  {'core': '[Ar]3d', 'rcut': [2.51, 2.49, 2.43], # test
        'vbar': ('f', 2.45), 'rcutcomp': 2.41,
        'filter': (0.63, 1.78),
        'empty_states': '5p'},
 'Sb': {'core': '[Kr]',   'rcut': [2.18, 2.33, 2.26], # test
        'vbar': ('f', 2.21), 'rcutcomp': 2.18,
        'filter': (0.61, 1.80)},
 'Hf': {'core': '[Kr]4d4f', 'rcut': [2.64, 2.47, 2.52], # test
        'vbar': ('f', 2.47), 'rcutcomp': 2.45,
        'filter': (0.63, 1.68),
        'empty_states': '6p'},
 'Re': {'core': '[Kr]4d5s4f', 'rcut': [2.51, 2.60, 2.47], # test
        'vbar': ('f', 2.50), 'rcutcomp': 2.50,
        'filter': (0.55, 1.73),
        'empty_states': '6p'},
 'Hg': {'core': '[Kr]4d4f5s', 'rcut': [2.47, 2.44, 2.46], # test
        'vbar': ('f', 2.36), 'rcutcomp': 2.43,
        'filter': (0.60, 1.65),
        'empty_states': '6p'},
 'Tl': {'core': '[Xe]4f', 'rcut': [2.29, 2.40, 2.25], # test
        'vbar': ('f', 2.30), 'rcutcomp': 2.25,
        'filter': (0.63, 1.66)},

You need also to edit (in the same file):
 'Hf': (72,
         ...
         (4, 2, 10, -7.6766379999999996),
         (4, 3, 14, -0.87157399999999996),
         (5, 0, 2, -2.0498280000000002),
         ....

 'Re': (75,
         ...
         (4, 2, 10, -9.5168160000000004),
         (5, 0, 2, -2.567348),
         (4, 3, 14, -1.9250799999999999),
        ...
 'Hg': (80,
         ...
         (4, 2, 10, -13.019221),
         (4, 3, 14, -4.1102910000000001),
         (5, 0, 2, -3.423486),
         ...

Note that you need to use eigensolver='cg' for setups that include 
semicore p-states,
and currently no lcao basis sets can be generated for them 
(https://trac.fysik.dtu.dk/projects/gpaw/ticket/187) (as well as for 
those containing semicore s).
Please let me know if you find any problems when testing - in such a 
case I would like
to add the python scripts showing the problem to 
https://svn.fysik.dtu.dk/projects/gpaw/trunk/gpaw/test/big/setups 
(currently empty).

Best regards,

Marcin

Jason Gruber wrote:
> Marcin,
> Thank you.  Yes, I'd be interested in trying those additional setup parameters.
> Jason
>
> On Wed, Feb 23, 2011 at 4:06 AM, Marcin Dulak <Marcin.Dulak at fysik.dtu.dk> wrote:
>   
>> Hi,
>>
>> Jason Gruber wrote:
>>     
>>> Hi all,
>>>
>>> Two miscellaneous questions to satisfy my own curiousity, i.e. they
>>> are not immediate problems.
>>>
>>> 1.  In the "Finding lattice constants / BCC iron" tutorial, I notice
>>> the code listed uses a "MethfesselPaxton" class that apparently isn't
>>> in GPAW, or at least not in the versions I've used.  Was there ever
>>> such a class?  Will there ever be one?  Any reasoning behind why or
>>> why not?
>>>
>>>       
>> this is available in the development version:
>> https://wiki.fysik.dtu.dk/gpaw/devel/developer_installation.html#developer-installation
>>
>>     
>>> 2.  There are a handful of elements for which there aren't setups
>>> listed on the "Atomic PAW Setups" page (now that I think of it, I
>>> should check the tarball also).  Is there any reason that some are
>>> missing (Tc is obvious - who cares?, but Y shows up a lot in high
>>> temp. superconductors).  Just wondering if there are some difficulties
>>> with these elements or that no one has bothered yet.
>>>
>>>       
>> i have experimental versions for all setups, but no enough tests (i
>> performed only bulk properties and eggbox/energy convergence tests) to
>> release them.
>> I can provide you with parameters to generate the setups.
>>
>> Marcin
>>     
>>> Jason
>>> _______________________________________________
>>> gpaw-users mailing list
>>> gpaw-users at listserv.fysik.dtu.dk
>>> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users
>>>
>>>
>>>       
>> --
>> ***********************************
>>
>> Marcin Dulak
>> Technical University of Denmark
>> Department of Physics
>> Building 307, Room 229
>> DK-2800 Kongens Lyngby
>> Denmark
>> Tel.: (+45) 4525 3157
>> Fax.: (+45) 4593 2399
>> email: Marcin.Dulak at fysik.dtu.dk
>>
>> ***********************************
>>
>>
>>     
>
>   

-- 
***********************************
 
Marcin Dulak
Technical University of Denmark
Department of Physics
Building 307, Room 229
DK-2800 Kongens Lyngby
Denmark
Tel.: (+45) 4525 3157
Fax.: (+45) 4593 2399
email: Marcin.Dulak at fysik.dtu.dk

***********************************