bayesian-methods

tfd.Categorical.quantile not working

Summary:

It'd be nice to have a builder pattern for var contexts to make them easy to construct for testing. Something that could be used like this:

MatrixXd m(3, 2); 
...
var_context vc
  = var_context::builder()
    .matrix("a", m)
    .real("f", 2.3)
    .build();

Current Version:

v2.23.0

The current example on MDN from Edward tutorials needs small modifications to run on edward2. Documentation covering these modifications will be appreciated.

Be conservative about the changes necessary to the core of DAPPER.

ATM, this requires pretending variable-length state or obs are fixed-length. This enables using np.ndarrays without hassle, but includes overhead.

• Is "pretending" compatible with all DA methods?
• How should RMSEs be calculated?

Suppose one chain is stuck on one mode; another on another mode. If those two chains may sample independently from each mode, the ESSs will be high when, really, they should be near zero since the samples don't represent anything like independently from the overall distribution. This is why multichain ESS makes sense and we should implement it. I feel like this will give a much better picture of t

This blogpost from Lyndon White mentions several antipatterns for Julia code: https://white.ucc.asn.au/2020/04/19/Julia-Antipatterns.html (thanks @bauglir for pointing this out). Some of the antipatterns mentioned here are also present in the FL code.

1. The most prominent one is the over-constraining of argument types. Some very specific constraints are needed for the update rules, but in oth