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
When I do inference with BayesNets.jl, sometimes the inference results for some attributes are NaNs instead of expected numeric values. I wonder if I misuse the BayesNets.jl or if this is a bug. Thanks for help!
Hello SEICS. Some of your variables have a zero likelihoods, ex Categorical([0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]). It is generally not recommended to allow an outcome to have zero likelihood unless you are 100% certain that value actually cannot ever be set. I suspect that is what is causing this issue.
You can check this by adding, say 1e-5, to all Categorical entries and re-normalizing.
Hi,
When I do inference with BayesNets.jl, sometimes the inference results for some attributes are NaNs instead of expected numeric values. I wonder if I misuse the BayesNets.jl or if this is a bug. Thanks for help!
Here is the example code I run:
The text was updated successfully, but these errors were encountered: