Add evidence computation tutorials

[z7076]

Add evidence computation based on harmonics with additional model.py to avoid multiprocessing problems.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[radka-j]

Thank you so much for adding this, it's looking great! Couple of initial thoughts.

1. We should add the log prob calculation as a function to AutoEmulate so that the user can just import. It should probably take the mcmc object as input.
   • Related to this, I was wondering whether it's also possible to update the mcmc implementation to save the log probs during the MCMC run but this is perhaps secondary. Have we looked into this?
2. We should also think about whether we want to add a wrapper class that does the harmonic calculation for the user (essentially wrapping the last cell in the notebook), this could also take in the mcmc object and call the log prob function (once implemented).
   • This should also add the harmonic package to the dependencies.
3. Once this functionality is in place, we should put together a tutorial that does a model comparison using the SIR and SEIR epidemic simulators (added in Add SEIR epidemic simulator #941) as discussed.