quantlop 0.0.4

High-performance quantum simulator for matrix-free Hamiltonian evolution

quantlop

High-performance quantum simulator for matrix-free Hamiltonian evolution

Introduction

quantlop is a high-performance simulator for the time evolution of quantum systems whose Hamiltonians can be written as sparse sums of Pauli words.

Rather than constructing the full Hamiltonian matrix, quantlop applies each Pauli word as a linear operator directly to the state vector. It then uses a Krylov method to numerically approximate the action of the matrix exponential. This matrix-free approach dramatically reduces memory usage and avoids costly dense-matrix operations, making larger simulations more practical.

Installation

Install the latest release of the package directly from PyPI with:

pip install quantlop

Quick example

Here is a simple code example using quantlop native data structures:

import numpy as np
import quantlop as ql

num_qubits = 3

# define Hamiltonian in Pauli basis
pwords = [
    ql.PauliWord(coeff=0.5, string="ZZI"),
    ql.PauliWord(coeff=0.2, string="YIX"),
]
ham = ql.Hamiltonian(pwords=pwords)

# set initial state vector
psi = np.zeros(2**num_qubits, dtype=complex)
psi[0] = 1.0

# evolve state vector
evolved_psi = ql.evolve(ham, psi)

The library also provides class methods to import Hamiltonians directly from other quantum computing frameworks:

• ql.Hamiltonian.from_pennylane to build from PennyLane Hamiltonian objects
• ql.Hamiltonian.from_qiskit to build from Qiskit SparsePauliOp objects

Multi-threading

Evolution is serial by default. Set num_threads to a positive integer to use that many OpenMP threads, or to "auto" to use the CPU count reported by the operating system:

evolved_psi = ql.evolve(ham, psi, num_threads="auto")

Development

The Python package is built with scikit-build-core, while the numerical C++ code is kept in the standalone quantlop_core CMake target.

Build project from source in dev mode and run Python tests with:

python -m pip install -e .[dev]
python -m pytest -v

Build the native C++ target from source with:

cmake -S . -B build
cmake --build build

Install pre-commit hook to format Python and C++ code automatically:

pre-commit install

Check the examples in the API docstrings and build the documentation locally with:

python -m sphinx -b doctest -W docs site
python -m sphinx -W docs site