Inference for DMLIV

Hi,

I am sorry to raise the issue.
The problem is that in the documentation I read that the DMLIV method' fit supports inference = 'bootstrap', but when I perform:

cate = DMLIV(model_Y_X(), model_T_X(), model_T_XZ(), 
             dmliv_model_effect(), dmliv_featurizer(),
             n_splits=N_SPLITS, # number of splits to use for cross-fitting
                      )

cate.fit(Y, T[:,0], X, Z, store_final=True, inference = 'bootstrap')

an error occurs:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-400-ac251344ad08> in <module>
      1 #Per ora devo mettere T[:,0] perchè accetta solo mono dim treatments
----> 2 cate.fit(Y, T[:,0], X, Z, store_final=True, inference='bootstrap')

TypeError: fit() got an unexpected keyword argument 'inference'

is there a way to make inference on such estimates?

Thank you,

Federico

Start at the DMLIV.fit entry point and reproduce the shown call with inference='bootstrap'. Compare the accepted fit arguments with the documentation claim about bootstrap inference. Done when the supported behavior and documentation agree, with the reported error either resolved or clearly documented.