Log Probability Implementation

[naseweisssss]

No description provided.

[coveralls]

Pull Request Test Coverage Report for Build 12766075584

Details

• 0 of 75 (0.0%) changed or added relevant lines in 1 file are covered.
• 37 unchanged lines in 1 file lost coverage.
• Overall coverage decreased (-2.7%) to 70.707%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/experimental/ProbabilisticGraphicalModels/bayesnet.jl	0	75	0.0%

Files with Coverage Reduction	New Missed Lines	%
src/experimental/ProbabilisticGraphicalModels/bayesnet.jl	37	0.0%

Totals
Change from base Build 12334000793:	-2.7%
Covered Lines:	1441
Relevant Lines:	2038

---

💛 - Coveralls

[naseweisssss]

This is the non-dp version for implementation, dp version will be addressed in another PR, still thinking of how to write the test for this

[sunxd3]

First, some high-level stuff. Julia has some conventions on Documentation, sadly, I don't think ChatGPT and Claude understand this yet.

I saw you removed deterministic functions, they do make implementation more difficult, but we can't get rid of them totally if we want to use this for future integration with JuliaBUGS' frontend.

How about try to emulate the implementation here to compute the log density, but adapt it to BayesianNetwork.

Dynamic Programming version is in no hurry, we first get the naive recursion correct, then we worry about make DP work.