qdd 0.3.0

Qiskit Provider for QDD

QDD

About

QDD is a decision diagram based quantum computing simulator. Memory usage can be significantly reduced compared to typical state vector based simulators.

It works as a Qiskit backend. At this point, we do not distribute wheel file, so you need to build it locally.

License

BSD 3-Clause Clear License

Installation

Standard (without MPI)

pip install qdd

With MPI support

pip install qdd-mpi

See README_MPI.md for prerequisites and details.

---

Supported environment

• Linux x86_64 with Python3.9 ~ 3.12

Install from source

Build

You must prepare the following software.

• cmake (>=3.25)
  • You can download the executable from Official GitHub.
  • If you use Ubuntu, the Kitware repository is easier to install.
• build
  • Just install it via pip.

You can manually build QDD as follows.

$ cmake . -DCMAKE_BUILD_TYPE=Release
$ cmake --build . -j
$ python3 -m pip install build
$ python3 -m build

BUILD_TYPE can be either Release, Debug or RelWithDebInfo. You can find an installable wheel file in 'dist' directory.

Alternatively, you can just execute the shell script in the project root to build it in one line:

$ ./scripts/local_build_and_test.sh build

If virtualenv is not used, it will create a new one named .venv-qdd under the project root. BUILD_TYPE is set to Release this way. If you want to just build C extension module (.so file), skipping distribution building process, you can specify cmake as the first argument of the shell script.

Test (Optional)

This source distribution provides functionality and performance tests with pytest and ./test/qdd_test.

Command

$ cmake . -DCMAKE_BUILD_TYPE=Release
$ cmake --build . -j
$ pip install -e .[test]
$ pytest test
$ ./test/qdd_test

If you want just to use this package, not editing itself, it is NOT recommended to execute these commands in your in-use virtualenv because it will install qdd package as editable, and qdd will not be installed under site-packages and python will try to import it from this directory whenever it is imported. In that case, it would be better to install qdd to your working virtualenv from binary distribution or without -e option, or execute these commands in a separated virtualenv. However, if you want to contribute to this package, it would be a nice way to keep using the environment because your edits will be reflected immediately to the package.

Alternatively, you can just execute the shell script in the project root to test this package in one line:

$ ./scripts/local_build_and_test.sh test

If virtualenv is not used, it will create a new one named .venv-qdd under the project root. This command includes the build of C++ extension module. BUILD_TYPE is set to Release.

If you pass all tests, please go on to create wheel file to install it as written in the section "Build" and "Wheel Installation". Note that some performance tests might fail depending on your computer's spec.

Wheel Installation

Please move to your working directory and install the wheel file created in the build phase.

$ pip install {QDD_DIR}/dist/qdd-XXX.whl

Usage

QDD works as a Qiskit backend.

from qiskit import QuantumCircuit
from qiskit.primitives import BackendSampler

from qdd import QddProvider

backend = QddProvider().get_backend()
circ = QuantumCircuit(3)
circ.h(0)
circ.cx(0,1)
circ.measure_all()
sampler = BackendSampler(backend=backend)
qdd_job = sampler.run(circuits=circ)
print(qdd_job.result())

QDD has its own implementation of the Sampler class

from qdd.qdd_sampler import Sampler
sampler = Sampler()
qdd_job = sampler.run(circuits=circ)
print(qdd_job.result())

If you need statevector, create the backend as follows.

backend = QddProvider().get_backend('statevector_simulator')

Limitation of Liability

In no event and under no legal theory, whether in tort (including negligence), contract, or otherwise, unless required by applicable law (such as deliberate and grossly negligent acts) or agreed to in writing, shall any Contributor be liable to You for damages, including any direct, indirect, special, incidental, or consequential damages of any character arising as a result of this License or out of the use or inability to use the Work (including but not limited to damages for loss of goodwill, work stoppage, computer failure or malfunction, or any and all other commercial damages or losses), even if such Contributor has been advised of the possibility of such damages.