I was not able to replicate correctly the main plots of Partial R2.
Ussing lasso
Using Random Forest
The main problem is based on the equivalent function I am using. So according to R we may code:
Only using model and treatment arguments.
But when I use PySensemakr I am forced to use the benchmark_covariates argument:
As you can see in the example above basically I forced using the intercept variable, but this is not correct.
I have checked the functions from this package and by construction, they forced users to use that argument, but please @anzonyqr could you please check the functions? and see whether we can do something to solve it?
I was not able to replicate correctly the main plots of Partial R2.
The main problem is based on the equivalent function I am using. So according to R we may code:
Only using model and treatment arguments. But when I use PySensemakr I am forced to use the benchmark_covariates argument:
As you can see in the example above basically I forced using the intercept variable, but this is not correct.
I have checked the functions from this package and by construction, they forced users to use that argument, but please @anzonyqr could you please check the functions? and see whether we can do something to solve it?