The ABINIT Users Mailing List ( CLOSED )

["Matteo Giantomassi" <Matteo.Giantomassi@uclouvain.be>]

> Dear Matthieu,
>
> On 18 Dec 2008, at 01:14, matthieu verstraete wrote:
>
>>
>> GW / PAW is not available in abinit 5.6
>
> GW/PAW is available in 5.6, see the release notes
> (http://www.abinit.org/Infos_v5.6/release_notes/release_notes.html B1)
> + automatic test v5#66 .
> However, it is true that GW/PAW is a relatively new feature, so it
> might be better in v5.7, soon to come ...
>


 Dear all,

 likely the problem encountered by Guangfu is due to the fact that, in
 case of PAW, the KSS file must be geratated using kssform==3.
 The direct diagonalization of the Hamiltonian (kssform==1, default method)
 is not yet implemented.

 There should be an explicative error message in the log file reporting
 that one has to use kssform==3.
 Please have a look at v5/Input/t66.in to understand how
 to use PAW+GW for a standard one-shot run.

 Anyway, as Xavier wrote, PAW+GW is still under development and
 there are features that are not tested or not coded.
 For example, in 5.6.x, self-consistent GW is not fully implemented,
 also the COHSEX approximation and ppmodel>=2 are not available.
 In 5.7 I've also decided to change
 the file format of the KSS file in case of PAW,
 thus "old" KSS files produced with 5.6.x won't be readable by 5.7

 If you only need the standard one-shot GW approach, 5.6.x should be OK
 For more involved calculations requiring self-consistency both
 on the eigenvalues and eigenfunctions, I suggest you wait for the
 release of 5.7

 Best Regards
 Matteo Giantomassi

> Xavier
>
>>
>>
>>
>> On Wed, Dec 17, 2008 at 6:48 PM, <6671011@163.com> wrote:
>> Dear friends,
>>
>>    I tried to run the tests file "tgw1_1.in" with PAW, but it did
>> not work.
>> (I added "pawecutdg", changed the value of "iscf", and replaced the
>> pseudopotential.)
>>
>>    After some simple tests, I found that the parameter "nbandkss"
>> conflicts
>> with the PAW method in the first step of GW calculation---when I
>> commented
>> "nbandkss", the first DATASET worked, when I uncommented it, abinit
>> exported
>> "leave_new : decision taken to exit ..."
>>
>>
>>  Is this true? Thanks for your help.
>>
>>
>>  p.s.  All the automatic tests have been successfully executed, so
>> the program
>> itself is probably OK.
>>
>>
>> Sincerely,
>> Guangfu Luo
>>
>>
>>
>> --
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Dr. Matthieu Verstraete
>>
>> European Theoretical Spectroscopy Facility (ETSF)
>> Dpto. Fisica de Materiales,
>> U. del Pais Vasco,
>> Centro Joxe Mari Korta, Av. de Tolosa, 72,   Phone: +34-943018393
>> E-20018 Donostia-San Sebastian, Spain        Fax  : +34-943018390
>>
>> Mail : matthieu.jean.verstraete@gmail.com
>> http://www-users.york.ac.uk/~mjv500
>
>