Fix bug in rabi 2D fit

[andrea-pasquale]

Closes #991.

I would like to add a test to verify that the code works properly.
I don't know if this is feasible unless we use some regressions.

Checklist:

• Reviewers confirm new code works as expected.
• Tests are passing.
• Coverage does not decrease.
• Documentation is updated.
• Compatibility with Qibo modules (Please edit this section if the current pull request is not compatible with the following branches).
  • Qibo: master
  • Qibolab: 0.1
  • Qibolab_platforms_qrc: main

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.50%. Comparing base (c77f563) to head (3377b87).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #997      +/-   ##
==========================================
- Coverage   97.56%   97.50%   -0.06%     
==========================================
  Files         123      123              
  Lines        9714     9714              
==========================================
- Hits         9477     9472       -5     
- Misses        237      242       +5     

Flag	Coverage Δ
unittests	97.50% <100.00%> (-0.06%)	⬇️

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

Files with missing lines	Coverage Δ
src/qibocal/protocols/rabi/amplitude_frequency.py	97.91% <100.00%> (ø)
...bocal/protocols/rabi/amplitude_frequency_signal.py	98.36% <100.00%> (ø)
src/qibocal/protocols/rabi/length_frequency.py	97.93% <100.00%> (ø)
.../qibocal/protocols/rabi/length_frequency_signal.py	98.37% <100.00%> (-1.63%)	⬇️

... and 1 file with indirect coverage changes

[alecandido]

I would like to add a test to verify that the code works properly.
I don't know if this is feasible unless we use some regressions.

I believe the best approach (in this and other cases) is truly a closure test: generate the data manually, as you expect they should be, with little or no noise (to make the fit simpler) and check that the fit is able to determine the parameters used in the data generation.

[andrea-pasquale]

Thanks for the approvals. Before merging I want to add a test following @alecandido's suggestion.

[alecandido]

Thanks for the approvals. Before merging I want to add a test following @alecandido's suggestion.

That's a correct approach, and I should encourage it.

However, if this is a bug fix, it may be more practical to just this PR as it is, and lift the suggestion to an issue, such that the test can be introduced as soon as there will be resources available to do that (i.e. decoupling it from the urgency of the bug fix itself).

[Edoardo-Pedicillo]

I have opened #1000 regarding the test issue, so we can merge this PR