[Q-e-developers] Fwd: possible commit in exx.f90

[Lorenzo Paulatto]

It looks correct to me, there are a few of these around the code 
already. But I did not test it!

cheers

On 07/24/2014 05:03 PM, Paolo Giannozzi wrote:
> Marco Govoni proposes the following patch to reduce
> the computational cost of hybrid functionals, by not
> computing terms in the exchange that have zero weight.
> Does anybody see any problem with it?
>
> Paolo
>
> On Thu, 2014-07-24 at 14:43 +0000, Marco Govoni wrote:
>> Caro Paolo,
>>
>> Pensi si possa mettere la seguente linea nel file exx.f90 in modo da ciclare
>> su jbnd ed ikk quando il peso wg e’ zero?
>> Se wg e’ zero, il totalizzatore dell’energia non dovrebbe incrementare.
>> In questo modo quando nbnd contiene stati vuoti, vengono saltati.
>>
>> Marco
>>
>> Index: exx.f90
>> ===================================================================
>> --- exx.f90	(revision 11099)
>> +++ exx.f90	(working copy)
>> @@ -1797,6 +1797,8 @@
>>          JBND_LOOP : &
>>          DO jbnd = 1, nbnd     !for each band of psi (the k cycle is outside band)
>>             !
>> +          IF ( ABS(wg(jbnd,ikk)) < eps_occ) CYCLE
>> +          !
>>             IF (noncolin) THEN
>>                 temppsic_nc = ( 0._dp, 0._dp )
>>             ELSE
>>
>>
>> --

-- 
Dr. Lorenzo Paulatto
IdR @ IMPMC -- CNRS & Université Paris 6
+33 (0)1 44 275 084 / skype: paulatz
http://www.impmc.upmc.fr/~paulatto/
23-24/4é16 Boîte courrier 115, 4 place Jussieu 75252 Paris Cédex 05