Censored data: v3 -> v4

[drbenvincent]

see #126

• Updates the notebook from v3 to v4 for the hackathon.
• Partial use of best practice.
• Changed the imputed model code - no longer uses pm.Bound
• Changed the unimpeded model code - uses pm.Censored

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2022-05-31T18:53:35Z
----------------------------------------------------------------

nitpick, maybe we can round to one or two digits

drbenvincent commented on 2022-05-31T19:41:55Z
----------------------------------------------------------------

done :)

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2022-05-31T18:53:36Z
----------------------------------------------------------------

This model makes use of PyMC's bounded variables.

drbenvincent commented on 2022-05-31T19:42:15Z
----------------------------------------------------------------

Well spotted. I've removed this now.

[drbenvincent]

done :)

---

View entire conversation on ReviewNB

[drbenvincent]

Well spotted. I've removed this now.

---

View entire conversation on ReviewNB

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

ricardoV94 commented on 2022-05-31T20:42:07Z
----------------------------------------------------------------

The dictionary warning should not be present with the latest aesara dependency from main

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

ricardoV94 commented on 2022-05-31T20:42:08Z
----------------------------------------------------------------

Perhaps import Interval on a distinct line so it's not so verbose in the distribution call?

from pymc.distributions.transforms import Interval`

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

ricardoV94 commented on 2022-05-31T20:42:09Z
----------------------------------------------------------------

Maybe expand slightly? "W can make use of pm.Censored to marginalize over the unobserved censored values.