fxc matrix elements

Anything regarding the post-processing utility (e.g. excitonic wavefunction analysis) is dealt with in this forum.

Moderators: Davide Sangalli, andrea marini, Daniele Varsano

Post Reply
sivanr
Posts: 5
Joined: Mon Dec 02, 2013 9:31 am

fxc matrix elements

Post by sivanr » Mon Dec 02, 2013 9:37 am

Dear yambo developers,

I am a new user, and currently exploring your code- mainly for TDDFT applications.

I would like to observe the fxc matrix elements (within linear response), for example in your SiH4 tutorial.
Is this possible?

Many thanks, and best regards

Sivan Refaely-Abramson
PhD student
Weizmann Institute of Science, Israel.

User avatar
Daniele Varsano
Posts: 4231
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: fxc matrix elements

Post by Daniele Varsano » Mon Dec 02, 2013 10:06 am

Dear Sivan,
the Fxc matrix elements are not stored in file, otherwise you will end up with an HD full very soon.
The TDDFT can be solved either in reciprocal space and in KS basis set. Depending on what implementations you are using you will have Fxc_GG' or <n|fxc|n'> matrix elements. Assuming you are interested in the latter, you can print them adding write commands in

Code: Select all

src/bse/K.F
the fxc matrix elements are calculated in tddft_alda_R_space function defined in

Code: Select all

./src/tddft/tddft_alda_r_space.F
and called here in src/bse/K.F:

Code: Select all

         if (BS_K_is_ALDA) H_res_x=H_res_x+&
&                          tddft_alda_R_space(iq,(/ic1,ic2/),(/ik1,ik2,ik1,ik2/),&
&                                                (/iv1,iv2/),(/is1,is2,is1,is2/),1)
Hope it helps.

Best,

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

sivanr
Posts: 5
Joined: Mon Dec 02, 2013 9:31 am

Re: fxc matrix elements

Post by sivanr » Tue Dec 03, 2013 7:58 am

Dear Daniele,

Thank you for your quick reply!

Can you please also comment on how to print the first option, namely Fxc_GG'?

Sivan.

User avatar
Daniele Varsano
Posts: 4231
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: fxc matrix elements

Post by Daniele Varsano » Tue Dec 03, 2013 11:22 am

Dear Sivan,
the Dyson equation:
\chi=\chi^0+\chi^0(v+Fxc)\chi
is solved here:
./src/pol_function/X_s.F

and the tddft kernel is calculated here:
/src/tddft/tddft_alda_g_space.F

through the libxc libraries.

Cheers,

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

sivanr
Posts: 5
Joined: Mon Dec 02, 2013 9:31 am

Re: fxc matrix elements

Post by sivanr » Wed Dec 04, 2013 7:51 am

Thanks a lot!

Sivan

sivanr
Posts: 5
Joined: Mon Dec 02, 2013 9:31 am

Re: fxc matrix elements

Post by sivanr » Sun Mar 30, 2014 1:55 pm

Hello again,

I used to print the real-space fxc matrix elements following your instructions from above, namely from within the bse/K.F code.

However, ever since I am using the new 3.4.1 version, I am having troubles with applying it. In fact, it seems like the real-space tdlda calculation does not use the K.o binary at all.

Do the matrix-elements still being calculated using tddft_alda_r_space.F? Can you please advice me regarding where should I approach in order to write those elements?

Thanks a lot!
Sivan

User avatar
Daniele Varsano
Posts: 4231
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: fxc matrix elements

Post by Daniele Varsano » Mon Mar 31, 2014 9:45 am

Dear Sivan,
please fill your signature with your affiliation.

The call to the alda kernel is still there:
./src/K.F

Code: Select all

462          ! :::ALDA        (resonant):::
463          !
464          if (BS_K_is_ALDA) H_res_x=H_res_x+&
465 &                          tddft_alda_r_space(iq,(/ic1,ic2/),(/ik1,ik2,ik1,ik2/),&
466 &                                                (/iv1,iv2/),(/is1,is2,is1,is2/),1)
467          !
468          ! :::ALDA        (coupling):::
469          !
470          if (BS_K_is_ALDA) H_cpl_x=H_cpl_x+&
471 &                          tddft_alda_r_space(iq,(/ic1,ic2/),(/ik1,ik2,ik1,ik2/),&
472 &                                                (/iv1,iv2/),(/is1,is2,is1,is2/),2)
473          !
you should add here your writing lines and recompile all.
Assure to do a make clean_all before reconfiguring and recompile the code.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

Post Reply