Enhancements to DRAG calibration

[andrea-pasquale]

Improvements for DRAG protocols.

• adding flips to DRAG http://login.qrccluster.com:9000/HOBOSa5ORAWry5k-xRaRww==/
• add drag protocol from Yale http://login.qrccluster.com:9000/oIhf1_PNRsmFzVi9EbfWsA==/
  - [ ] add detuning
• update documentation

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 96.98%. Comparing base (9723091) to head (2faf087).

Additional details and impacted files

@@            Coverage Diff             @@
##              0.2    #1055      +/-   ##
==========================================
+ Coverage   96.96%   96.98%   +0.02%     
==========================================
  Files          98       99       +1     
  Lines        7932     8039     +107     
==========================================
+ Hits         7691     7797     +106     
- Misses        241      242       +1     

Flag	Coverage Δ
unittests	96.98% <100.00%> (+0.02%)	⬆️

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

Files with missing lines	Coverage Δ
src/qibocal/protocols/__init__.py	100.00% <100.00%> (ø)
src/qibocal/protocols/drag.py	100.00% <100.00%> (ø)
src/qibocal/protocols/drag_simple.py	100.00% <100.00%> (ø)

... and 2 files with indirect coverage changes

[andrea-pasquale]

@hay-k I'm asking a review from you since we had some discussions in the past about the drag calibration which was failing from time to time in your recalibration script. I think that using the nflips variable should reduce the chances of the fit failing since as you can see http://login.qrccluster.com:9000/0VD0ziRdS1SGw2NhY-Gv3g==/. Eventually you can give it a try also with the other protocol drag_simple.

[hay-k]

I am also glad to see, that you have clarified the definition of beta, and the division by anharmonicity is not happening anymore. Back when I was looking into the DRAG code, I found it really confusing. It looked like part of the code was following a notation where beta is the entire coefficient in front of the quadrature component, and other parts of code were following a notation where the coefficient in front of the quadrature component is beta/anharmonicity.

[andrea-pasquale]

I am also glad to see, that you have clarified the definition of beta, and the division by anharmonicity is not happening anymore. Back when I was looking into the DRAG code, I found it really confusing. It looked like part of the code was following a notation where beta is the entire coefficient in front of the quadrature component, and other parts of code were following a notation where the coefficient in front of the quadrature component is beta/anharmonicity.

Yes, we already made that change in #990. Some people follow the notation of beta/anharmonicity. In our case we don't really care.