y0-causal-inference/y0

Error in DSL.py in building Probability objects from strings applying conditional logic to sets

rjc955 opened this issue · 2 comments

rjc955 commented

Typing P({X}|{Y,Z}) should create a Probability object with the representation P(X|Y,Z). Instead, it creates P(X,Y,Z).

Conditional_Probability_from_Sets_Issue
cthoyt commented

Nope, this is totally correct. You're using the bar operator between two sets, which takes the union of the sets. Then the union set XYZ is passed to the construction which makes a joint distributor over all of them

  • 👍1
rjc955 commented

Got it, thanks very much!