Introduce crosstalk_matrix as Qubit attribute

[andrea-pasquale]

Required by qiboteam/qibocal#684.

When it will get merged I will also update accordingly qibolab_platforms.

Checklist:

• Reviewers confirm new code works as expected.
• Tests are passing.
• Coverage does not decrease.
• Documentation is updated.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (402f550) 63.03% compared to head (493526e) 63.02%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #735      +/-   ##
==========================================
- Coverage   63.03%   63.02%   -0.02%     
==========================================
  Files          48       48              
  Lines        6623     6620       -3     
==========================================
- Hits         4175     4172       -3     
  Misses       2448     2448              

Flag	Coverage Δ
unittests	63.02% <100.00%> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[alecandido]

Are you sure it should go under each single qubit? It should be a matrix in the first place, rotating from all qubits (a weight for each one) to all qubits.

Wouldn't it be more appropriate to put it under a qubits container?
It could be the QubitMap, but it's potentially better to keep that simple (id: Qubit) and hold this information directly in the Platform.

[andrea-pasquale]

Are you sure it should go under each single qubit? It should be a matrix in the first place, rotating from all qubits (a weight for each one) to all qubits.

Wouldn't it be more appropriate to put it under a qubits container? It could be the QubitMap, but it's potentially better to keep that simple (id: Qubit) and hold this information directly in the Platform.

The current approach is the first thing that came to my mind. Indeed, it should a be a matrix. I am not really sure about putting it under a qubits container. Mainly because I don't want to overcomplicate the runcard if necessary. A possible solution could be to put a row of the matrix under each qubit (similarly to what I am doing with the only difference that instead of storing it as a dict we store directly as an array).

Also regardless of how we store it, we can add a method to the platform which generates directly the crosstalk matrix.

[alecandido]

Also regardless of how we store it, we can add a method to the platform which generates directly the crosstalk matrix.

But it's going to be used altogether anyhow, and it makes connections between multiple qubits. If there is going to be a method in the platform, we could well put the whole attribute there.

[stavros11]

Following the meeting, we decided to merge this now and refactor later (if needed) when implementing the mechanism for flux corrections.