IR intensities in psi4 1.5


Is it now possible to get IR intensities?

I’ve read that it was a goal to implement this for 1.4 and then for 1.5. And I saw on GitHub that there might be code for this in the developer version (psi4/ at master · psi4/psi4 · GitHub, line e.g 417).

I tested to calculate the frequencies for one example using frequency(“b3lyp-d3bj”) using psi4 1.5, but besides the frequencies I did not seem to get any intensities.

Do I need to call or add something else in order to get the intensity information?

  • We have intensities for analytic frequencies. This is currently only available for closed-shell HF and LDA functionals, so not for B3LYP.
  • Intensities for frequencies by finite difference of analytic gradients are theoretically possible. It’s something we’ve wanted ever since 1.4, but it kept getting pushed back. This feature will be added for either the 1.6 release next month or the 1.7 release in November/December. I’ll have a better estimate on which of the two it will be, after Wednesday.
  • Intensities by finite difference of energies are not theoretically possible.
1 Like

Ok, thank you for this information. I keep my fingers crossed that it will be added in 1.6! :slightly_smiling_face: It would be very nice of you, if you could write here after Wednesday when the intensity calculation for frequencies by finite difference of analytic gradients is planned to be added.

The hope is to get this in (alongside the larger driver refactor that we promised in the 1.4 paper) for 1.6, to be released towards the end of May.

Obligatory disclaimer that nothing is certain until the code is merged in. But we’ve begun merging in pieces of the driver refactor, so we’re optimistic about this target.

I’ll post here if plans changed, or this particular piece of the driver refactor gets merged in.

Thank you for the update. Great to hear that you are optimistic that it will be part of 1.6.