Keith Battocchi

We've published a new version of our package that should be compatible with the latest sklearn releases. See if `pip install -U econml` fixes it for you.

@Sandy4321 No, we typically use synthetic data because even if we had real data, without somehow knowing the ground truth causal effect there would be no way to demonstrate that...

If I understand your question correctly, the interpretation of the treatment effect shouldn't change, but the optimization problem changes slightly when there is more than one treatment, which could result...

You should set `discrete_treatment=True` in this case, though it may not matter much in practice. When discrete treatment is specified, we one-hot-encode the treatment and then drop the first column,...

DML does not have a way to directly model a discrete outcome, so our first-stage fitting logic just calls `predict` on whatever Y model you have passed in and then...

This is expected; as you note the `ate_` attribute applies a double-robustness correction to the computation of the ATE itself (on the training data); the `ate()` method allows you to...

Thanks for pointing this out! I think we can now safely drop support for sklearn

Yes, exactly, all of the DML methods estimate partially linear models. But note that: 1. It is linear in the treatment T, but CausalForestDML estimates a flexible conditional average treatment...

To make the example realistic, we use a mis-specified model (where the functional form of what we're estimating does not exactly match the data-generating process). We take the log transform...

And just to be even more explicit - the true treatment effect really is gamma(X)*T/Y (this is not an approximation); the issue with our mis-specified model is that we are...