Error

[ZSCDumin]

learner_x = BaseRRegressor(XGBRegressor(), control_name='control')
learner_x.fit(X=x_train, treatment=treatment_train, y=y_train)
cate_x_no_p = learner_x.predict(X=x_test)

Occur error when using BaseRRegressor, how to handle this problem?

[ThatShelbs]

Having the same issue. Can run the other Base Regressors just fine but when I run BaseRRegressor I get the same error.

[ppstacy]

Hi @ZSCDumin and @ThatShelbs, Thanks for reporting this!

Do you mind provide the full logs for this error? It's a bit difficult now to find out where cause this issue. From the code it didn't provide the propensity score input, so R Learner will automatically run a propensity model to generate the propensity scores. It looks like this error coming from here.

Could you provide the data which can reproduce this error that I can take a quick look? Thanks!

[ThatShelbs]

causalml_baserlearner_issue

The notebook of interest is basic_causalml.ipynb. Included the notebook that generates synthetic data just in case it is needed.

[jeongyoonlee]

Thanks, @ThatShelbs for sharing the notebook. The issue has been fixed in PR #170 and working notebook is available in gist. A new version with the fix will be published to PyPI in a few days.