It'd be helpful if I could spell

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.58%. Comparing base (cc604a4) to head (68b40de).

@@            Coverage Diff             @@
##           master     #797      +/-   ##
==========================================
+ Coverage   93.54%   93.58%   +0.03%     
==========================================
  Files          26       26              
  Lines        3519     3521       +2     
==========================================
+ Hits         3292     3295       +3     
+ Misses        227      226       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

This issue appears to be resolved.

But looking at the fix: is the result that the there's no longer cut sharing when there are some 0 probabilities? I would prefer that the code just inefficiently solved children with 0 probabilities than didn't share cuts.

(A user would always have the option to not include an arc with 0 probability if there's a preference for reducing the number of solves.)

Currently we share cuts iff they have the identical set of children. We could improve that to nodes with have a subset of children. So yes, this might result in a less efficient solve; probably somewhere in between setting refine_at_similar_nodes = false, depending on the structure of the zeros.

But yeah, I'll take another go at fixing this.

Okay, I'm taking a look at this.

The issue is

The fix of just adding this is surprisingly subtle, because risk measures now need to deal with values that have zero probability mass

WorstCase is the obvious one, but it already checks this:

Belief states broke. I assume because something weird happens if you try to update the belief from somewhere that it was impossible to come from. There's a subtle distinction between ambiguity sets and nodes with similar children...

Bump @adow031 interested in your thoughts