OFEX: OpenFermion EXpansion Toolkit

---

Overview

OFEX (OpenFermion EXpansion) is a Python toolkit that builds upon OpenFermion to provide tools for classical quantum simulation, measurement optimization, efficient classical data structures for quantum objects and quantum algorithms like QKSD (Quantum Krylov Subspace Diagonalization).

It is modular, flexible, and focused on reducing quantum measurement costs and improving simulation performance.

---

Key Features

• Quantum State Tools: Manipulate quantum states, Fock states, and chemical reference states.
• Clifford Operations: Simulate Clifford-based transformations and standard operations.
• Measurement Optimization: Perform Iterative Coefficient Splitting (ICS) and Killer Shift optimizations.
• Quantum Operation Utilities: Perform diagonalizations, expectation values, and operator applications.
• Pauli Hamiltonians: Construct chain, ring, and Heisenberg models.
• Quantum Krylov Subspace Diagonalization (QKSD): Tools for simulating Krylov subspaces and advanced sampling.

---

Installation

To install OFEX, clone this repository and install the dependencies:

git clone https://github.com/snow0369/ofex.git
cd ofex
pip install -r requirements.txt

We recommend to install psi4 with 'Installer' option in here. After then, add environment variable PSI4PATH for example:

# ~/.bashrc
...
export PSI4PATH=/home/username/path/to/psi4conda/bin/psi4
...

Also, replace openfermionpyscf/_pyscf_molecular_data.py with this version.

Dependencies

• Python 3.8+
• NumPy
• SciPy
• OpenFermion
• Refer to requirements.txt for further dependencies.

---

Getting Started

1. Expectation Value of Sparse Operators

Operation with sparse and dense states.

import numpy as np
from openfermion import QubitOperator

from ofex.linalg.sparse_tools import expectation
from ofex.state import BinaryFockVector, int_to_fock
from ofex.state.state_tools import state_allclose, pretty_print_state

# Define operators
z0, z1 = QubitOperator("Z0"), QubitOperator("Z1")

# Define states  |01>
state_0_dense = np.array([0, 0, 1, 0])  # A state can be np array
state_0_sparse = {BinaryFockVector((1, 0)): 1}  # or dict of (BinaryFockVector, complex)
state_0_int_to_fock = {int_to_fock(2, num_qubits=2): 1}  # Binary Fock vector can be initialized by an integer
print(state_allclose(state_0_dense, state_0_sparse)) # True
print(state_allclose(state_0_dense, state_0_int_to_fock)) # True  - They are all close to each other.
print(pretty_print_state(state_0_dense)) # +1 |01>

# Define a state  |10>
state_1_sparse = {BinaryFockVector((0, 1)): 1}

# Compute expectation value
print(expectation(z0, state_0_sparse))  # -1.0
print(expectation(z1, state_0_sparse))  #  1.0
print(expectation(z0, state_1_sparse))  #  1.0
print(expectation(z1, state_1_sparse))  # -1.0

2. Clifford Diagonalization

2. Iterative Coefficient Splitting (ICS)

4. Quantum Krylov Subspace Diagonalization (QKSD)

---

API Overview

Full API Documentation

| Module | Description | |---------------------------------|---------------------------------------------------------------------------------------------| | ofex.clifford | Clifford operations and tools (tableaus, Pauli diagonalization). | | ofex.hamiltonian | Hamiltonian examples (chain, ring, Heisenberg, PPP-Polyacene). | | ofex.linalg.sparse_tools | Sparse matrix tools: expectation, diagonalization, transitional amplitudes, etc. | | ofex.measurement | Measurement optimization (Sorted Insertion, Iterative Coefficient Splitting, Killer Shift). | | ofex.propagator | Exact and Trotter-based propagators, with both of real and imaginary time evolutions. | | ofex.sampling_simulation | Sampling-based simulation tools, including Hadamard tests. | | ofex.state | Quantum state tools, including binary Fock states. | | ofex.transforms | Fermion-to-Qubit transformations and Fermion rotations and factorization. | | ofex_algorithms.qksd | QKSD algorithms and simulation tools. |

---

Contributing

We are not yet prepared to accept contributions at this time.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contact

For issues, feedback, or contributions, reach out via the repository's Issues page.

---

Acknowledgments

This project builds upon the powerful OpenFermion library and numerical tools provided by NumPy and SciPy.