GLM poisson regression update with bambi

[chiral-carbon]

Addresses issue #86 and updates GLM poisson regression to use bambi instead of pymc.glm

---

Still pretty new to bambi so wanted to understand how to get mu in the posterior group in the resulting inference data on running bambi.Model.fit() in cell 19.

[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 2021-06-29T09:43:15Z
----------------------------------------------------------------

Line #6.    import patsy as pt

Instead of patsy you could use formuale. This is what bambi uses.

from formulae import design_matrices

chiral-carbon commented on 2021-06-29T13:30:32Z
----------------------------------------------------------------

oh okay sure, will use this

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2021-06-29T09:43:16Z
----------------------------------------------------------------

dm = design_matrices(fml, df, na_action="error")

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2021-06-29T09:43:16Z
----------------------------------------------------------------

dm.common.as_dataframe()

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2021-06-29T09:43:17Z
----------------------------------------------------------------

The names used by formulae are slightly different

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2021-06-29T09:43:17Z
----------------------------------------------------------------

mu is not present

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

aloctavodia commented on 2021-06-29T09:43:18Z
----------------------------------------------------------------

mu is not present.

Maybe instead you can compute the posterior predictive distribution and then take the mean. Something like this.

model.posterior_predictive(inf_fish_alt) #notice tthat by default this is an inplace operation

inf_fish_alt.posterior_predictive["nsneeze"].mean().item()

chiral-carbon commented on 2021-06-29T13:32:31Z
----------------------------------------------------------------

yes wanted some help as to how to get mu with bambi. thanks a lot!

[chiral-carbon]

oh okay sure, will use this

---

View entire conversation on ReviewNB

[chiral-carbon]

yes wanted some help as to how to get mu with bambi. thanks a lot!

---

View entire conversation on ReviewNB

[chiral-carbon]

@aloctavodia hey! sorry I took so long to update this, I had been taking a break from the internship for personal reasons. hope you and @OriolAbril can review this when you get some time.

[aloctavodia]

thanks @chiral-carbon!