search

dscolby / CausalELM.jl

Taking causal inference to the extreme!
https://dscolby.github.io/CausalELM.jl/
MIT License
21 stars 0 forks source link

Add W argument #39

Closed dscolby closed 3 months ago

dscolby commented 4 months ago

Add W parameter for covariates for the treatment function in double machine learning, X learning, doubly robust estimation, and R learning.

dscolby commented 3 months ago

This seems kind of unnecessary. The original paper doesn't use a W parameter. EconML does this so let you include a bunch of characteristics in a model and only estimate the CATE conditional on a subset of those variables. Hence, the W is the set of all potential confounders and the X are the variables of interest. I think we can close this out (at least for now).

dscolby commented 3 months ago

Completed with commit b84bb08.