Fix causal impact calculation error in synthetic control and interrupted time series

[drbenvincent]

• There was an error in the causal impact calculations which affected Interrupted Time Series and Synthetic control. The causal impact was being calculated using the expectation of the posterior predictive distribution. But it should have been based on the actual posterior predictive values which have additional likelihood-related observation noise. Therefore, previous estimates of causal impact would have slightly underestimated the uncertainty.
• I also re-ran the affected notebooks
• Closes Ensure causal impact is calculated from posterior predicted observations, not expected values #294

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 77.06%. Comparing base (d5e493c) to head (7541516).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #296   +/-   ##
=======================================
  Coverage   77.06%   77.06%           
=======================================
  Files          21       21           
  Lines        1378     1378           
=======================================
  Hits         1062     1062           
  Misses        316      316           

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

[drbenvincent]

I'll cut a point release when this is merged.

[juanitorduz]

Good catch!