qpe-toolbox 1.1.0

A Quantum Phase Estimation Toolbox

qpe-toolbox is an open-source Python package for compiling and simulating Quantum Phase Estimation (QPE) circuits, combining quantum chemistry with tensor-network methods. The code is hosted on github, and docs are available on GitHub Pages.

See our overview presentation for the motivation and philosophy behind the toolbox.

Built on the tensor-network library quimb, it provides:

• Classical preprocessing: quantum chemistry with PySCF and fermionic encodings via OpenFermion
• Quantum simulation: QPE circuit construction and circuit-level simulator with tensor networks
• Postprocessing: energy retrieval from phase measurement outcomes

Installation

Requirements

Our package is built above numpy/scipy/matplotlib, openfermion, pyscf and quimb as core dependencies. We use jax for variational circuit optimization. jupyterlab and jupytext are needed to run the examples as notebooks. The complete list of dependencies is in pyproject.toml.

Installation from pypi

qpe-toolbox is available on pypi. Install it with

pip install qpe-toolbox

Installation from sources

Installing from sources gives access to our tutorials which contains detailed explanations on the Quantum Phase Estimation algorithm.

with pip

# clone the project
git clone [e-mail]:quobly-sw/qpe-toolbox.git && cd qpe-toolbox

# Create a virtual environment
python3 -m venv .venv --prompt qpe-toolbox

# activate it
source .venv/bin/activate

# install the package and its dependencies
pip install -e .[dev]

with uv

# clone the project
git clone [e-mail]:quobly-sw/qpe-toolbox.git && cd qpe-toolbox

# Create a virtual environment and synchronize it with lock
uv sync --locked

# activate it
source .venv/bin/activate

Contents

The package is divided in four modules:

• circuit: creation and manipulation of quimb circuits.
• hamiltonian: class for defining Hamiltonians and interface with pyscf for chemistry.
• estimation: perform different flavors of Quantum Phase Estimation.
• tensor: manipulation of Matrix Product Operators (MPO) and Matrix Product States (MPS).

Examples

We provide a list of notebooks that introduce the basics of the package and contain detailed explanations on the Quantum Phase Estimation algorithm. They are available in the examples directory as plain .py files using the py:percent format. We use Jupytext to convert them and pair them with a twin .ipynb notebook. To convert a given example and execute it as a notebook, open jupyterlab, right-click on the .py file and select "Open with > Notebook" or "Jupytext Notebook": this allows you to save the notebook's outputs in your local repository. We also include the executed notebooks in our documentation.

1. building_circuits explains how to create, plot, record and load quantum circuits in quimb and qiskit.

2. chemistry_to_qubit describes how to build the qubit Hamiltonian and perform the Density Matrix Renormalization Group (DMRG) algorithm for a given molecule.

3. textbook_qpe introduces the textbook Quantum Phase Estimation algorithm assuming time evolution is implemented exactly.

4. trotter_decomposition introduces the Trotter-Suzuki decomposition to implement a time evolution operator $U$.

5. qpe_with_trotter studies the Quantum Phase Estimation algorithm using Trotterization of the evolution operator, and provides resource estimates.

6. qpe_with_lcu gives an introduction to Block Encoding via Linear Combination of Unitaries.

7. robust_phase_estimation introduces the Robust Phase Estimation algorithm, based on the Hadamard test circuit.

8. performance_mps compares the performance of quimb and qiskit when contracting and sampling circuits with Matrix Product States.

9. hyperoptimization presents advanced contraction schemes provided by quimb.

10. variational_circuit_preparation finds an initial guess state with variational circuit optimization.

Basic workflow

To perform Quantum Phase Estimation with the toolbox, take the following steps:

1. The Hamiltonian class describes the qubit Hamiltonian. Choose a system:

   • Spin model with e.g. heisenberg_hamiltonian or a custom Hamiltonian instance.
   • Molecule with chemistry_hamiltonian; see the chemistry_to_qubit example.

2. Prepare an initial state as a Matrix Product State. Two methods are available:

   • Density Matrix Renormalization Group (DMRG) - see the chemistry_to_qubit tutorial.
   • Parametrized circuit optimization - see the tutorial on variational_circuit_preparation.

3. Encode the Hamiltonian into a unitary via either:

   • Exact time evolution or Trotterization, available as methods of the Hamiltonian class - see the tutorial on trotter_decomposition.
   • Block encoding functions from the estimation module - see the tutorial on Linear Combination of Unitaries: qpe_with_lcu.

4. Initialize a circuit with a physical register and a phase register. From the circuit module, chose between:

   • make_circ to create a Tensor Network representation of the circuit.
   • make_circMPS to store the state as an MPS and iteratively apply the gates.

   See the tutorials on building_circuits, performance_mps and hyperoptimization for an introduction on circuit simulation with quimb.

5. Run QPE: in the estimation module, choose between

   • Textbook QPE: see the corresponding tutorial textbook_qpe.
   • Robust Phase Estimation, a version of QPE with a single ancilla and circuit repetitions - see the robust_phase_estimation tutorial.

Contributing

Contributions are welcome and highly appreciated. To get started, check out the contributing guidelines.

License

Apache 2.0

Credits

2026 Foxconn, Quobly

Authors:

• Thibaud Louvet ([e-mail])
• Calvin Ku ([e-mail])
• Yu-Cheng Chen ([e-mail])
• Carlos Ramos Marimón ([e-mail])
• Olivier Gauthé ([e-mail])
• Tristan Meunier ([e-mail])
• Min-Hsiu Hsieh ([e-mail])
• Benoit Vermersch ([e-mail])