Quantum computing library (qclib)

Qclib is a quantum computing library implemented using qiskit.

The focus of qclib is on preparing quantum states, but it is not limited to that.

Instalation

The easiest way of installing qclib is by using pip:

pip install qclib

Initializing your first quantum state with qclib

Now that qclib is installed, you can start building quantum circuits to prepare quantum states. Here is a basic example:

$ python

import numpy as np
from qiskit import transpile
from qiskit.providers.aer.backends import AerSimulator
from qclib.state_preparation import LowRankInitialize

# Generate 3-qubit random input state vector
n = 3
rnd = np.random.RandomState(42)
input_vector = rnd.rand(2 ** n) + rnd.rand(2 ** n) * 1j
input_vector = input_vector / np.linalg.norm(input_vector)

# Build a quantum circuit to initialize the input vector
circuit = LowRankInitialize(input_vector).definition

# Construct an ideal simulator
backend = AerSimulator()

# Tests whether the produced state vector is equal to the input vector.
t_circuit = transpile(circuit, backend)
t_circuit.save_statevector()
state_vector = backend.run(t_circuit).result().get_statevector()
print('Equal:', np.allclose(state_vector, input_vector))
# Equal: True

Comparing algorithms for quantum state preparation

The following table shows the depth, number of qubits, and CNOTs of circuits produced by different state preparation methods for the same random 15-qubit state.

method	lib	qubits	cnots	depth
low-rank	qclib	15	30999	53645
svd	qclib	15	38813	71581
isometry	qiskit	15	32752	131011
multiplexor	qiskit	15	65504	131025
bdsp	qclib	1151	72320	1603
dcsp	qclib	32767	262016	899

You can reproduce the results here.

Authors

The first version of qclib was developed at Centro de Informática - UFPE. Qclib is an active project, and other people have contributed.

If you are doing research using qclib, please cite our project. We use a CITATION.cff file, so you can easily copy the citation information from the repository landing page.

License

qclib is free and open source, released under the Apache License, Version 2.0.