Vasilis/drate #391
Vasilis/drate #391
Conversation
…ith sklearn logic for future easier extendibility.
…ded a summary functionality to the causal forest.
| raise AttributeError("`oob_predict` is not available when " | ||
| "the estimator was fitted with `warm_start=True`") |
There was a problem hiding this comment.
Should this be mentioned in the method docstring? What about the warm_start part of the init docstring?
There was a problem hiding this comment.
Added it in docstring. Also added it in the warm_strt docstring. I also removed warm_Start from causalForestDML. It makes no sense there as there is no way for the final model to be refit to new data! So that only is a weird option to have.
econml/dml/causal_forest.py
Outdated
| if not self._drate: | ||
| raise AttributeError("Doubly Robust ATE calculation on training data " | ||
| "is available only when `drate=True`!") | ||
| return self._ate |
There was a problem hiding this comment.
For some reason the coverage report shows this line (and the corresponding ate_stderr_ one as uncovered... Does that make sense?
econml/dml/causal_forest.py
Outdated
| over the training data and with a doubly robust correction. | ||
| Available only when `discrete_treatment=True` and `oob=True`. | ||
| """ | ||
| return NormalInferenceResults(d_t=self._d_t[0] if self._d_t else 1, |
There was a problem hiding this comment.
This and the subsequent properties are also showing as uncovered by tests.
There was a problem hiding this comment.
This is weird! This should be covered!
…ring of method and docstring of warm_start
No description provided.