Integrate Quac with Qrochet

[Todorbsc]

Closes #7

[codecov]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (482ac90) 56.98% compared to head (65bc6bc) 60.76%.
Report is 1 commits behind head on master.

❗ Current head 65bc6bc differs from pull request most recent head c033992. Consider uploading reports for the commit c033992 to get more accurate results

Files	Patch %	Lines
ext/QrochetQuacExt.jl	86.95%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master       #9      +/-   ##
==========================================
+ Coverage   56.98%   60.76%   +3.77%     
==========================================
  Files           5        6       +1     
  Lines         186      209      +23     
==========================================
+ Hits          106      127      +21     
- Misses         80       82       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Todorbsc]

Three issues:

• Do we need to check if the gates' matrices (tensors) have the right values?
• Is there any case of a quantum circuit where Biyections are needed to correctly build the sites?
• Should we add a test with "Swap" quantum gates?

[mofeing]

Yes! This is the kind of tests I was referring to.

I think the "all inner indices are not sites" condition can be too strict, but is ok for now. What do you think @jofrevalles?

LGTM, but just suggested changes.

[jofrevalles]

I think the "all inner indices are not sites" condition can be too strict, but is ok for now. What do you think @jofrevalles?

Yes, I agree that it is too strict, but for now I think this test is OK. I would maybe add a test to see if the contraction of the TensorNetwork works and that's it.

[mofeing]

I would maybe add a test to see if the contraction of the TensorNetwork works and that's it.

It's better to not mix tests: That's a test for contract(::TensorNetwork), not for Quantum.

[mofeing]

• Do we need to check if the gates' matrices (tensors) have the right values?

No, that is the responsability of Quac.

• Is there any case of a quantum circuit where Biyections are needed to correctly build the sites?

I don't understand the question. Can you rephrase it?

• Should we add a test with "Swap" quantum gates?

I don't think we need a special test for that. Do you have an edge case in mind?

[Todorbsc]

• Do we need to check if the gates' matrices (tensors) have the right values?

No, that is the responsability of Quac.

Perfect!

• Is there any case of a quantum circuit where Biyections are needed to correctly build the sites?

I don't understand the question. Can you rephrase it?

In Tenet, Biyections were used to find the sites (which was called interlayer) so I thought maybe that was necessary (https://github.com/bsc-quantic/Tenet.jl/blob/v0.3.0/ext/TenetQuacExt.jl#L35)

• Should we add a test with "Swap" quantum gates?

I don't think we need a special test for that. Do you have an edge case in mind?

Well, I suggested a test for a quantum circuit with swap gates just to have a complete code coverage of the extension.

[mofeing]

In Tenet, Biyections were used to find the sites (which was called interlayer) so I thought maybe that was necessary (https://github.com/bsc-quantic/Tenet.jl/blob/v0.3.0/ext/TenetQuacExt.jl#L35)

Nope, that was another implementation.

Well, I suggested a test for a quantum circuit with swap gates just to have a complete code coverage of the extension.

Well, it's just another gate. Nothing special should happen. If you want to translate the Swap gate to a real swap of the tensor indices, then that is a Transformation and so, it's not the responsability of Quantum.