Issue with simulating categorical variables

[sagarsimha]

Hi,

There is a bug in simulating the categorical variables. Instead of simulating from the predicted probability distribution, the current code picks the class with the maximum probability. This was raised here originally in the R version. Kindly look into it.

CausalInference/gfoRmula#47 (comment)

[LilJing]

Hi, thanks for posting! The modification has been made locally, our released version will be updated soon

Best,
Jing

[LilJing]

Hi, a kind reminder that the bug has been fixed and is included in the released version 1.1.7. You can upgrade with: pip install --upgrade pygformula==1.1.7