Skip to content

Releases: unitaryfund/mitiq

v0.22.0

31 Jan 16:48
55481c7
Compare
Choose a tag to compare

Changelog

Summary

In this release we focused on improving our documentation with three new examples, new REM docs, and navigation improvements, along with some bug fixes. This release also adds a new module calibration which allows one to run a series of experiments to see what error mitigation parameters will work best for their particular setup. This feature is still under active development.

Many thanks to @amirebrahimi for his continued work on making readout error mitigation available, and usable in Mitiq.

Thanks to everyone who contributed to Mitiq this year! It's been a great time for error mitigation, and we look forward to continuing to grow Mitiq in the new year! Happy holidays! 🎄🎉🎊

All changes

v0.21.0

30 Nov 20:17
2b9a5db
Compare
Choose a tag to compare

Summary

This release officially adds support for the learning-based PEC sub-technique which is now fully documented and ready to be applied by Mitiq users. We are still assessing the stability of this new sub-technique, so if you notice any bugs, please let us know by opening issues on GitHub.

Functions to apply Readout Error Mitigation (REM) are also introduced in this release, special thanks to Amir Ebrahimi for this contribution!

Also, the noise scaling by identity insertion method is included in the ZNE section of the user guide. Special thanks to Purva Thakre for this contribution!

During the release cycle we accepted the RFC for implementation of calibration tools (Solution 1). We also completed a prototype of this approach, which will be released in a future version of Mitiq.

In addition, this release adds support for qubit-independent representations for PEC, along with bug fixes and minor dependency upgrades.

What's Changed

Full Changelog: v0.20.0...v0.21.0

v0.20.0

31 Oct 22:35
aed2b15
Compare
Choose a tag to compare

Summary

This milestone focused on updating our support for numpy (1.23), adding a tutorial demonstrating the learning-based PEC workflow, and scoping out what device/noise calibration might look like as part of Mitiq. Additionally identity insertion has been added as a noise-scaling technique available for mitigation protocols such as zero-noise extrapolation. Expect more documentation of this feature in future releases. Big thanks to @purva-thakre for getting this in Mitiq!

There are also some minor bug fixes, documentation updates, and a new example contributed by @nickdgardner as well!

All changes

v0.19.0

03 Oct 21:16
2d1b7b8
Compare
Choose a tag to compare

Changelog

Version 0.19.0

Summary

With this release Mitiq supports most recent versions of Python: 3.8, 3.9 and 3.10!
We drop support for Python 3.7.

Mitiq is now compatible with Numpy 1.21.6. Different versions of NumPy may not work properly.

Another important update is the addition of new tools for applying learning-based PEC!
This release introduces a function for learning depolarizing noise representations from Clifford circuit data.
Read more in our updated API-doc.

Special thanks to the external contributors @yitchen-tim, @amirebrahimi and @isaac-gs!

All Changes

v0.18.0

06 Sep 17:29
9695d2d
Compare
Choose a tag to compare

This release cycle focused on review and approval of two RFCs, one for Readout Error Mitigation (REM) #1387 and one for Identity insertion noise scaling #335 (not listed as PRs). It also includes bug fixes and minor dependency upgrades.

What's Changed

Full Changelog: v0.17.1...v0.18.0

Release v0.17.1

09 Aug 18:32
fc1c17c
Compare
Choose a tag to compare

Changelog

Version 0.17.1

Summary

This patch release includes support for the latest versions of Qiskit (0.37.1), Cirq (1.0.0), and Pyquil (3.2.1), along with other minor dependency upgrades and bug fixes.

All Changes

Release v0.17.0

06 Jul 23:20
da6142f
Compare
Choose a tag to compare

Changelog

Version 0.17.0 (July 6th, 2022)

Summary

This release includes contributions from UnitaryHACK 2022! 🎉 We had 3 merged pull requests, and a fourth is looking to be merged in the next milestone. Along with the great contributions from hackers, this release focused on expanding our ZNE examples to other frontends (Cirq, Braket, and Pennylane), building out learning-based PEC, and improving our benchmarking infrastructure.

All Changes

Release v0.16.0

03 Jun 19:56
759cba1
Compare
Choose a tag to compare

Changelog

Version 0.16.0 (June 3rd, 2022)

Summary

This release officially adds support for the digital dynamical decoupling (DDD) technique which is now fully documented and so ready to be applied by Mitiq users. This is still very new technique and so, if you notice any bugs, please let us know by
opening issues on GihHub.
A further notable addition is the function generate_quantum_volume_circuit() by @nickdgardner, extending the Mitiq benchmarking module with quantum volume
circuits.

Congratulations to the new member of the Mitiq team @natestemen and special thanks to the external contributors @Aaron-Robertson, @nickdgardner, @ZhaoyiLi-HekJukZaaiZyuJan and @amirebrahimi!

All Changes

Release v0.15.0

29 Apr 19:03
103969e
Compare
Choose a tag to compare

Changelog

Version 0.15.0 (April 29th, 2022)

Summary

This milestone focused on updating dependencies and making progress on two new features, dynamical decoupling and learning based PEC. For dynamical decoupling, high-level functions and rules were added. For learning-based PEC, a function calculating representations with a biased (combination of depolarizing and dephasing) noise model was added. Several high priority bugs and issues were also fixed.

Special thanks to new contributors @RubidgeCarrie and @nickdgardner for their contributions to this release!

All Changes

Release v0.14.0

06 Apr 14:40
612595f
Compare
Choose a tag to compare

Changelog

Version 0.14.0 (April 6th, 2022)

Summary

This milestone focused on updating dependencies and making progress on two new features, dynamical decoupling and learning based error mitigation techniques. A number of high priority bugs and issues alo were fixed.

All Changes