OpenFermion

1.7.1 verified Mon Apr 27 auth: no python

OpenFermion is an open-source library for quantum chemistry simulation on quantum computers. Version 1.7.1 (requires Python >=3.10) provides tools for mapping fermionic Hamiltonians to qubit operators, variational quantum eigensolver (VQE) workflows, and interfaces to quantum computing SDKs. Released periodically, approximately quarterly.

pip install openfermion

Common errors

error AttributeError: 'MolecularData' object has no attribute 'n_qubits' ↓

cause MolecularData was not loaded. Accessing attributes before `.load()` fails.

fix

Add molecule.load() after constructing the MolecularData object.

error ImportError: cannot import name 'InteractionOperator' from 'openfermion' ↓

cause In versions >=1.0, the import path changed. InteractionOperator is still available but must be imported from `openfermion.ops` only in older versions.

fix

Use from openfermion import InteractionOperator (v1.0+) or from openfermion.ops import InteractionOperator (v0.x). Check your version.

error ModuleNotFoundError: No module named 'openfermionpyscf' ↓

cause openfermionpyscf is a separate plugin package for PySCF integration. Not included by default.

fix

Install with pip install openfermionpyscf and import as import openfermionpyscf.

Warnings

breaking In v1.0, many top-level imports changed from submodules (e.g., `openfermion.ops`, `openfermion.utils`) to direct `openfermion`. Code using old import paths will break. ↓

fix Replace e.g. `from openfermion.ops import InteractionOperator` with `from openfermion import InteractionOperator`.

deprecated `get_ground_state` function is deprecated. Use `eigensolver` from `openfermion` or `scipy.sparse.linalg.eigsh`. ↓

fix Use `from openfermion import eigensolver` or `scipy.sparse.linalg.eigsh`.

gotcha When using `MolecularData`, the molecule must be loaded with `.load()` before accessing properties. Forgetting this raises an `AttributeError`. ↓

fix Always call `molecule.load()` after creating `MolecularData` object.

gotcha The Jordan-Wigner transform of a large Hamiltonian can produce a very dense operator; consider using Bravyi-Kitaev or other transforms for efficiency. ↓

fix Use `of.bravyi_kitaev(hamiltonian)` or `of.symmetric_jordan_wigner(hamiltonian)` to reduce qubit count or operator weight.

Imports

get_sparse_operator

wrong

from openfermion.ops import get_sparse_operator

correct

from openfermion import get_sparse_operator

Deprecated path; function is now at top-level

qubit_operator_sparse

wrong

from openfermion.utils import qubit_operator_sparse

correct

from openfermion import qubit_operator_sparse

Moved to top-level in v1.0+

MolecularData

wrong

from openfermion.hamiltonians import MolecularData

correct

from openfermion import MolecularData

Deprecated submodule; use top-level import

Quickstart

Basic workflow: load molecular data, obtain fermionic Hamiltonian, map to qubit operator, and inspect sparse matrix representation.

import openfermion as of

# Create a simple hydrogen molecule (minimal basis)
geometry = [('H', (0.0, 0.0, 0.0)), ('H', (0.0, 0.0, 0.7414))]
molecule = of.MolecularData(geometry, basis='sto-3g', multiplicity=1, charge=0)
molecule.load()

# Get the fermionic Hamiltonian
hamiltonian = molecule.get_molecular_hamiltonian()
print(hamiltonian)

# Map to qubit Hamiltonian (Jordan-Wigner)
qubit_ham = of.jordan_wigner(hamiltonian)
print(qubit_ham)

# Get sparse matrix
sparse_mat = of.get_sparse_operator(qubit_ham)
print(sparse_mat.shape)