Printing tuned parameters in CausalForestDML

[JFernandoGR]

Hi,

Is it possible to modify "tune" option to print optimized hyperparameters, or just adding another function to do it?

https://econml.azurewebsites.net/_modules/econml/dml/causal_forest.html#CausalForestDML.tune

For writing papers, it would be good to add it.

Thanks,

Jairo.

[kbattocchi]

This is a great suggestion. In the meantime, note that you can read the optimized hyperparameters out of the class itself (e.g. if you are using params='auto', then read off the min_weight_fraction_leaf, max_depth, and min_var_fraction_leaf attributes from the estimator after tuning and they'll give you the discovered values; if you're using an explicit parameter dictionary then read off the corresponding parameters instead).