Refactoring HAAQI to save precomputations

[groadabike]

closes #363

This is the first draft of the refactoring of HAAQI to precompute and reuse some parts of the code.
This logic can the be applied to HASPI and HASQI

Need to run a more extensive sanity test to ensure the consistency of the output compared with the old implementation

TODO:

• Run the metric several times to measure consistency with old implementation
• Create unit and regression tests

[groadabike]

@jonbarker68, this is a draft of this refactoring.
if you have a minute, can you take a look to see if makes sense to you?

[codecov]

Codecov Report

Attention: Patch coverage is 99.11243% with 6 lines in your changes are missing coverage. Please review.

Project coverage is 93.80%. Comparing base (d78922f) to head (6c9b9cb).
Report is 1 commits behind head on main.

❗ Current head 6c9b9cb differs from pull request most recent head 099ce34. Consider uploading reports for the commit 099ce34 to get more accurate results

Files	Patch %	Lines
clarity/evaluator/ha/pyhaaqi.py	98.14%	6 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #364      +/-   ##
==========================================
+ Coverage   92.88%   93.80%   +0.91%     
==========================================
  Files          43       46       +3     
  Lines        3921     4597     +676     
==========================================
+ Hits         3642     4312     +670     
- Misses        279      285       +6     

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

[groadabike]

The output of this refactoring is not exactly the same as previous implementation due to the different order in when random noises are generated.