v0.3.0
Summary
Factories now support "batched" executors, meaning that when a backend allows for the batch execution of a collection of quantum circuits, factories can now leverage that functionality. In addition, the main focus of this release was implementing probabilistic error cancellation (PEC), which was introduced in Temme2017 as a method for quantum error mitigation. We completed a first draft of the major components in the PEC workflow, and in the next release plan to demonstrate the full end-to-end operation of the new technique.
All Changes
- Fix broken links on the website (@erkska, gh-400).
- Use
cirq v0.9.0
instead ofcirq-unstable
(@karalekas, gh-402). - Update
mitiq.about()
(@rmlarose, gh-399). - Refresh the release process documentation (@karalekas, gh-392).
- Redesign factories, batch runs in
BatchedFactory
, fix Qiskit utils tests (@rmlarose, @andreamari, gh-381). - Add note on batched executors to docs (@rmlarose, gh-405).
- Added Tensorflow Quantum executor to docs (@k-m-schultz, gh-348).
- Fix a collection of small build & docs issues (@karalekas, gh-410).
- Add optimal QPR decomposition for depolarizing noise (@karalekas, gh-371).
- Add PEC basic implementation assuming a decomposition dictionary is given (@andreamari, gh-373).
- Make tensorflow requirements optional for docs (@karalekas, gh-417).
Thanks to @erkska and @k-m-schultz for their contributions to this release!