You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In the course of working on #240 and #236, @rfl-urbaniak ran into problems when attempting to use infer_discrete and TraceEnum_ELBO in conjunction with causation and explanation queries defined via chirho.counterfactual.ops.preempt. This issue tracks progress on diagnosing and fixing those problems.
The text was updated successfully, but these errors were encountered:
@rfl-urbaniak it's been a few weeks since you showed us this, and I'm afraid I've lost track of the context since then. Could you please create a failing test that illustrates this behavior and paste it here or push it to a branch in this repository and put a link to the branch here? Once you do that I will work on a fix.
@eb8680 A most recent version of a notebook illustrating what seems to be a problem with using enumeration with explanation-related handlers is now in the ac-archive (PR #403 ), in the enumeration-explanation.ipynbnotebook. If this gets resolved, the notebooks will need some updating, otherwise, I'm using sampling at least for now, and this shouldn't make much difference, as the number of possible combinations in the examples used in the notebooks is somewhat limited.
In the course of working on #240 and #236, @rfl-urbaniak ran into problems when attempting to use
infer_discrete
andTraceEnum_ELBO
in conjunction with causation and explanation queries defined viachirho.counterfactual.ops.preempt
. This issue tracks progress on diagnosing and fixing those problems.The text was updated successfully, but these errors were encountered: