2
$\begingroup$

Are there any electronic structure programs that print the exchange-correlation (XC) integrals for a given DFT functional? These integrals are of the form $$\langle pq|w|rs\rangle$$ where $w$ is an XC kernel specific to a given functional and $pqrs$ label arbitrary MOs.

I'm trying to make a Python code that can do TDHF/KS, with the prior SCF done using an existing electronic structure package. I have it working fine for TDHF, but TDKS requires these XC integrals and I can't find any program that prints them (or the atomic orbital basis equivalent). Ideally, if this option exists in Gaussian or Psi4, that would be most helpful to me, but I would be interested to see if any program will print these.

$\endgroup$
  • $\begingroup$ Just a guess, but psi or pyscf should at least be modifiable to print it, if it is not possible yet. $\endgroup$ – pH13 - Yet another Philipp Apr 26 at 17:42
  • 1
    $\begingroup$ @Martin-マーチン I've been using Gaussian output files to get other integrals. I've been looking around the "Interfacing to Gaussian" page on their website, but have yet to come across what I'm looking for. I will have to check out the CCL. $\endgroup$ – Tyberius Apr 26 at 18:14
  • 1
    $\begingroup$ It seems that the only integrals Gaussian can print is this one: IOp(3/33) Integral package printing. 0 No integrals are printed. 1 Print one-electron integrals. 3 Print two-electron integrals in standard format. 4 Print two-electron integrals in debug format. 5 Combination of 1 and 3. 6 Combination of 1 and 4. $\endgroup$ – OhLook May 8 at 3:18
  • 1
    $\begingroup$ @Martin-マーチン Why do you discourage people from using CCL? I've seen lots of useful answers on there. $\endgroup$ – OhLook May 8 at 3:19
  • 1
    $\begingroup$ @OhLook That's true, and I don't discourage anyone to use it (I use it). I don't usually recommend it here; when the question is very much on topic, it only doubles the effort (avoid cross posting) for everyone. On a more general note, I find the CCL very hard to search through and sometimes you only get a very quick reply (as it happened in this case, iirc someone said to contact Gaussian support). $\endgroup$ – Martin - マーチン May 8 at 6:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.