Installation

OpenQL is available from PyPI as a pre-built package for Windows, MacOS, and Linux for all active Python 3.x versions. Once you have Python and have access to a command line you can get it as follows:

pip install qutechopenql

after which you should be able to run

python -c 'import openql; print(openql.get_version())'

to see if it works.

Note

Depending on your OS and Python configuration, you may need to use python3 instead of just python to disambiguate with a Python 2.7 installation, and/or use pip3 or python -m pip instead of just pip. You may also need to add --user at the end of the pip command to avoid permission problems. If you’re unsure of what all of the above means, first read up on how Python and pip work on your operating system in the relevant Python (or Linux distribution) documentation; installation of Python packages is rather fundamental to Python and out of scope for this manual.

Warning

The documentation you’re reading now is generated for version 0.10.5. If there is a mismatch, be aware that there may be an API mismatch as well! The reference information can, however, be queried from within Python using the help() builtin and (from 0.8.1.dev6 onwards) using the various dump_*() functions.

If you’re on MacOS and want to use the visualizer, you’ll need XQuartz in addition. You can install this using Brew.

Some of OpenQL’s components are optional when OpenQL itself is compiled. In general, the pre-built package includes everything, except for initial placement due to a license conflict. If you need initial placement, you’ll need to compile manually.

OpenQL used to support Conda in addition to PyPI/pip for Python package management, but ultimately this was disabled due to excessive time spent on dependency resolution. Nevertheless, the Conda recipe is still available, so it may or may not work, but for this you’ll also have to compile manually, as the prebuilt Conda packages are likely out of date.