IQM Pulla

IQM Pulla#

Pulla (pulse-level access) is a client-side software which allows the user to control the generation and execution of pulse schedules on a quantum computer. Within the existing IQM QCCSW stack, PulLA is somewhere between circuit-level execution and EXA-experiment.

An interactive user guide is available as a Jupyter notebook in the docs folder.

Use#

Create a virtual environment and install dependencies:

conda create -y -n pulla python=3.11 pip=23.0
conda activate pulla
pip install iqm-pulla[notebook, qiskit, qir]

The [qiskit] option is to enable Qiskit-related features and utilities, like converting Qiskit circuits to Pulla circuits, constructing a compatible compiler instance, or constructing a PullaBackend for running Qiskit jobs.

The [qir] option is to enable QIR support, e.g. the qir_to_pulla function.

The [notebook] option is to be able to run the example notebooks, using and run it in Jupyter Notebook:

jupyter-notebook

Development#

Install development and testing dependencies:

pip install -e ".[dev,notebook,qiskit,qir,testing,docs]"

e2e testing is execution of all user guides (Jupyter notebooks). User guides cover the majority of user-level features, so we achieve two things: end-to-end-test Pulla as a client library, and make sure the user guides are correct. (Server-side use of Pulla is e2e-tested as part of CoCoS.)

You have to provide CoCoS and Station Control URLs as environment variables:

COCOS_URL=<COCOS_URL> STATION_CONTROL_URL=<SC_URL> tox -e e2e

Notebooks are executed using jupyter execute command. It does not print any output if there are no errors. If you want to run a particular notebook and see the output cells printed in the terminal, you can use nbconvert with jq (https://jqlang.github.io/jq/download/) like so:

jupyter nbconvert --to notebook --execute  docs/Quick\ Start.ipynb --stdout | jq -r '.cells[] | select(.outputs) | .outputs[] | select(.output_type == "stream") | .text[]'

Run unit tests, build docs, build package:

tox
tox -e docs
tox -e build