Refactoring Qibolab Emulator

[lballerio]

In this PR I worked on the following:

• Fixing emulator simulator: when I first started working on the emulator, I realized it was not propagating correctly the results from the solver engine (in this case QuTiP).
• Added Classical Crosstalk cause by Drive lines (their parameters are given in the hamiltonian definition).
• format some structures in Qubit and HamiltonianConfig classes, specifically moving transmon_levels from HamitlonianConfig to Qubit.
• Modified how AcquisitionMode and AveragingMode process data and added Confusion Matrix correction for the computed probabilities.
• Added first draft of documentation for the emulator.
• Adding crosstalk for Flux channels.
• Adding new tests for new emulator.
• Changing the logic for crosstalk factors for both drive and flux channels

Checklist:

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

[codecov]

Codecov Report

❌ Patch coverage is 90.18405% with 16 lines in your changes missing coverage. Please review.
✅ Project coverage is 39.71%. Comparing base (98b5b6a) to head (cdb2fcf).
⚠️ Report is 16 commits behind head on main.

Files with missing lines	Patch %	Lines
src/qibolab/_core/instruments/emulator/results.py	89.61%	8 Missing ⚠️
...qibolab/_core/instruments/emulator/hamiltonians.py	90.41%	7 Missing ⚠️
src/qibolab/_core/instruments/emulator/emulator.py	92.30%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1401      +/-   ##
==========================================
+ Coverage   38.96%   39.71%   +0.75%     
==========================================
  Files         120      120              
  Lines        6216     6320     +104     
==========================================
+ Hits         2422     2510      +88     
- Misses       3794     3810      +16     

Flag	Coverage Δ
unittests	39.71% <90.18%> (+0.75%)	⬆️

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

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[lballerio]

In 0c506d2 I decided to change _single_shot_results in results.py module.
In fact I decided to remove changes in #1509 and also delete the usage of inverse and direct mapping for the measurements.
The first decision was taken for the sake of simplycity of the code: I think implementing confusion matrices with those changes would be inconvenient.

Deleting np.unique usage was due to the following fact:
We decided to insert this in order to keep correlation between simultaneous measurements, i.e. if qubits 1 and 2 were measured at the same time T, the shots had to be the same, so both have to measure state 4;
however this condition does not hold anymore when adding confusion matrices and the correlation is lost because of the error in the measurement. This is why I decided to take the full probability tensor and multiply by the tensor product of all the qubits' confusion matrices and after that sample.
I am open to discussions on the topic and new suggestions.

[lballerio]

I encourage reviewers to give me feedback on the 'tensorization' of the functions _cyclic_results and _singleshot_results; if you don't understand or you have simpler and better ideas please let's discuss!