How to perfom DFT calculation with IntegralSolverPySCF?

[NanaEngo]

Is there any way to perform DFT calculation with IntegralSolverPySCF?

[alexfleury-sb]

Hello @NanaEngo, this feature is not yet implemented. IntegralSolverPySCF is meant to compute one- and two-body integrals from a mean-field approximation, to eventually generate a second-quantized Hamiltonian. Since DFT deals with electronic density, it would require another design for IntegralSolver, which would work in the first quantization framework.

There are some work aiming at mapping DFT equations on quantum device, such as Senjean et al. (2022) and Ko et al. (2023). However, if your goal is to solely compare energies / densities with a classical DFT solution, it is possible to use directly PySCF (here I am assuming PySCF is installed as your question refers to IntegralSolverPySCF).

[NanaEngo]

Thank you for your prompt reply.
Precisely, it is the VQE calculation from the DFT that interests me. And my hope was to take advantage of the different modules already existing in Tangelo

[alexfleury-sb]

Hello again @NanaEngo, are you basing your idea on a published paper? If not, what would be the envisioned workflow?

[NanaEngo]

Hello Alexandre, Your idea is base on 1. https://arxiv.org/abs/2302.03052 2. http://arxiv.org/abs/2312.12386v3 And we want to calculate Linear response properties of some OLED molecules (large organic molecules). *Nana Engo, PhD* Director of Students Affairs University of Buea

…

On Mon, 26 Aug 2024 at 15:04, Alexandre Fleury ***@***.***> wrote: Hello again @NanaEngo <https://github.com/NanaEngo>, are you basing your idea on a published paper? If not, what would be the envisioned workflow? — Reply to this email directly, view it on GitHub <#407 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALM6GYHLHXSDBFW4YSHZSTDZTMYW5AVCNFSM6AAAAABM77VP26VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTANBVGI2DAMY> . You are receiving this because you were mentioned.Message ID: ***@***.*** com>

[alexfleury-sb]

After a quick glance at the papers, I can say the SA_OO_Solver would be useful to compute excited with orbital optimization in your case (as shown in the second paper).

For the Hamiltonian definition in a VQE-in-DFT framework as described in the first reference, there is no complete workflow to do that in Tangelo. I suggest using directly PySCF as described in an earlier response to get integrals or a fermionic Hamiltonian, and then convert it into a tensor product of Pauli words with our toolset.