Connectivity constraints for clifford synthesis

[JakobSchaeffeler]

Description

This PR implements connectivity constraints for the clifford synthesis as requested by #328, by passing optional coupling maps. The passed coupling map used to restrict CNOTs to only connected qubits. Additionally, the clifford synthesizer needs to check if the circuit can be synthesized with any permutation of qubits. If no coupling map is passed all to all connectivity is assumed.

Checklist:

• The pull request only contains commits that are related to it.
• I have added appropriate tests and documentation.
• I have made sure that all CI jobs on GitHub pass.
• The pull request introduces no new warnings and follows the project's style guidelines.

[codecov]

Codecov Report

Attention: Patch coverage is 95.87629% with 8 lines in your changes missing coverage. Please review.

Project coverage is 90.3%. Comparing base (1170b3d) to head (e189af9).

Files with missing lines	Patch %	Lines
include/cliffordsynthesis/CliffordSynthesizer.hpp	82.3%	3 Missing ⚠️
include/cliffordsynthesis/Results.hpp	81.2%	3 Missing ⚠️
src/cliffordsynthesis/CliffordSynthesizer.cpp	75.0%	1 Missing ⚠️
src/cliffordsynthesis/Tableau.cpp	98.1%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #506     +/-   ##
=======================================
- Coverage   90.4%   90.3%   -0.1%     
=======================================
  Files         87      87             
  Lines      10052   10199    +147     
  Branches    1722    1759     +37     
=======================================
+ Hits        9093    9217    +124     
- Misses       959     982     +23     

Flag	Coverage Δ
cpp	90.0% <95.5%> (-0.2%)	⬇️
python	96.8% <100.0%> (+0.8%)	⬆️

Files with missing lines	Coverage Δ
include/cliffordsynthesis/Tableau.hpp	86.6% <ø> (ø)
include/cliffordsynthesis/encoding/GateEncoder.hpp	90.0% <100.0%> (+1.1%)	⬆️
include/cliffordsynthesis/encoding/SATEncoder.hpp	100.0% <ø> (ø)
...lude/cliffordsynthesis/encoding/TableauEncoder.hpp	100.0% <ø> (ø)
src/cliffordsynthesis/encoding/GateEncoder.cpp	100.0% <100.0%> (ø)
...rc/cliffordsynthesis/encoding/MultiGateEncoder.cpp	100.0% <100.0%> (ø)
src/cliffordsynthesis/encoding/SATEncoder.cpp	83.7% <100.0%> (+0.3%)	⬆️
...c/cliffordsynthesis/encoding/SingleGateEncoder.cpp	100.0% <100.0%> (ø)
src/cliffordsynthesis/encoding/TableauEncoder.cpp	91.0% <100.0%> (+4.2%)	⬆️
src/mqt/qmap/clifford_synthesis.py	100.0% <100.0%> (+5.2%)	⬆️
... and 4 more

... and 4 files with indirect coverage changes

[burgholzer]

Note

#507 just went in and brought some updates from the MQT Core dependency that cause some conflicts here. Could you please resolve these? They should be fairly easy to resolve. Sorry for the inconvenience.

[JakobSchaeffeler]

@burgholzer I tried to resolve the conflicts and everything works locally. However the CI/Change/Check test fails, but all subtasks in there are marked as successful. Do you know what the issue is?

[burgholzer]

@burgholzer I tried to resolve the conflicts and everything works locally. However the CI/Change/Check test fails, but all subtasks in there are marked as successful. Do you know what the issue is?

Thanks for fixing the conflicts so quickly.
I just retriggered the respective CI job. Let's see if it still fails on the second try.
Should be unrelated to your changes in any case.

[pehamTom]

Very nice work!

I actually do not have too much to comment on this. The only thing that could be nicer from a maintainability standpoint is to get rid of the offsets in the encodings of every constraint. That might require some more involved rewrites. Unless you have a quick and elegant idea to get rid of this, we can leave it for future refactoring.

One additional feature that would go a long way is to also include a small example for this functionality in the documentation. Extending the Jupyter Notebook for https://mqt.readthedocs.io/projects/qmap/en/latest/Synthesis.html by a small example should be sufficient.

[JakobSchaeffeler]

@burgholzer @pehamTom I incorporated the feedback by Tom. However now all Python Test fail with a Deprecation warning and one C test which should be unrelated to my PR. Do you know how to solve that?

[pehamTom]

@burgholzer @pehamTom I incorporated the feedback by Tom. However now all Python Test fail with a Deprecation warning and one C test which should be unrelated to my PR. Do you know how to solve that?

Yeah, that is not your fault. We leave the PR open until the IDP is over anyway. I will take care of it.

[burgholzer]

@burgholzer @pehamTom I incorporated the feedback by Tom. However now all Python Test fail with a Deprecation warning and one C test which should be unrelated to my PR. Do you know how to solve that?

I think if you hit the "Update Branch" button (effectively merging in the main branch), the deprecation warnings should be gone. IIRC, I fixed them for QMAP in one of the latest updates.
Otherwise, I'll probably address them next week in a separate PR. They should not be related to your changes here 😌

[burgholzer]

Just merged the main branch in which contains the fixes from #509.
The deprecation warnings here are now really related to your code changes and should be addressed (should be fairly easy).