Qiskit OpenFermion

[PabloAMC]

Abstract

OpenFermion is the leading library to perform quantum chemistry calculations, including calculation of Hamiltonians integrals in different basis, mapping to qubits or generation of circuits for the implementation of chemistry algorithms. OpenFermion currently has integration with chemistry packages with Psi4 and has support of Cirq (Google), Forest (Rigetti) and OpenBoson (Xanadu), but it is open source so having support of Qiskit would be useful to users of Qiskit.

Description

This project would probably require the implementation of code that allows the implementation of Hamiltonian simulation circuits in Qiskit. OpenFermion-Cirq also allows for the creation of Variational Circuits and Ansatzes, Trotter Hamiltonian simulation methods, and several gates.

Members

• @githubhandle
• @githubhandle - Slack: @slackhandle email: [email protected]
• Qiskit Coach: @githubhandle

Deliverable

A small integration library with qiskit aqua

GitHub repo

[HuangJunye]

Qiskit has its own chemistry module but I don't know how well it compare to OpenFermion. I think Steve Wood (@woodsp-ibm) would be the right person to consult.

Hi Steve, we started this repo very recently for Qiskit advocates to propose project ideas. It would be great if we can get your expert opinions on this project.

[obliviateandsurrender]

I'm not sure how helpful I can be, but would surely want to contribute to this project.

[woodsp-ibm]

Hi I know OpenFermion has some finer grained building block such as ladder operations for building out operators etc. Its something that is down as an issue in Qiskit Chemistry for such a feature, but something that is not there at present. Is the thought to have some bridge/adapter type capability to say build out operators, ansatz circuits etc in OpenFermion and then be able to use/import/convert these such that they can be used within Qiskit Chemistry/Aqua. For instance a native Qiskit QuantumCircuit can be given say to VQE as an ansatz however that circuit came about.

[HuangJunye]

Thanks @woodsp-ibm for your reply! I really not familiar with Qiskit Chemistry so I won't even attempt to make a comment :P @PabloAMC What do you think about Steve's reply?

[woodsp-ibm]

@pbark FYI and in case you have any thoughts/input

[danielbultrini]

I'm actually quite keen on this! I think I can start working on it in a couple of weeks as I am currently busy - definitely a translation layer is the way to go, perhaps through QASM?