search

BAMresearch / bayem

Implementation and derivation of "Variational Bayesian inference for a nonlinear forward model." [Chappell et al. 2008] for arbitrary, user-defined model errors.
MIT License
2 stars 1 forks source link

Separate VB algorithm #39

Closed TTitscher closed 3 years ago

TTitscher commented 3 years ago

It seems that the focus of this package shifts from being focused on (aka built around) the Chappell VB implementation towards our idea of "convenience code" to define somewhat general inference problems. Nothing against that!

Should we, for clarity, then separate our VB implementation? Especially, if the "convenience code" is about to be moved elsewhere.

joergfunger commented 3 years ago

I would rather try to extract the relevant features of our implementation into taralli, only afterwards I would then think about adapting our package again. In the optimal setting, this would then be two additional sub packages in taralli.