dowhy
dowhy copied to clipboard
CATE estimation conditioned on a confounding variable?
Based on some tutorials, for example :https://microsoft.github.io/dowhy/example_notebooks/dowhy-conditional-treatment-effects.html It seems to me that CATE is defined w.r.t effect modifiers. The effect modifiers are the ones that affect only to the outcome, eg. X0, X1 variables in the tutorial.
Can we compute CATE conditioned a confounding variable. For example, age variable can affect to the treatment and outcome, and we want to compute the CATE given a specific age range.
Thanks
yes, you can do that. Just add the variable age
as both a common_cause and as a effect modifier.