qiskit-noise-learning

A Python toolkit for randomization-based quantum noise characterization.

Installation

You can install qiskit-noise-learning via pip from the GitHub repository (main branch):

pip install "git+https://github.com/Qiskit/qiskit-noise-learning@main"

For visualization support, include the visualization dependencies:

pip install "qiskit-noise-learning[vis] @ git+https://github.com/Qiskit/qiskit-noise-learning@main"

See the contribution guidelines for developer dependencies and editable installations.

Interfaces and Documentation

This library has two levels of interface. The first is the low-level interface where the user directly interacts with objects representing core concepts in noise learning, enabling custom design of every aspect of a noise learning protocol. The second is a higher-level interface, the qiskit_noise_learning.protocols subpackage, which holds one pair of functions per protocol — prepare_vanilla_program and process_vanilla_result for the standard "vanilla" protocol — each pair wrapping a stock workflow into a single call on either side of execution.

See the following tutorials for examples:

• docs/tutorials/learning_protocol.md — end-to-end use of the protocol functions
• docs/tutorials/workflow.md — step-by-step walkthrough of the internal pipeline

Development

python -m pytest          # run tests (includes doctests)
ruff check --fix . && ruff format .   # lint and format
pre-commit run --all-files            # all checks

See the contribution guidelines for details on developer setup, testing, building the documentation, and the release notes workflow.

License

Apache 2.0

Citing this package

If you use this package in your research, use the CITATION.bib file in this project’s repository to cite the appropriate reference(s).