Adam Li

Some API issues to figure out. In sklearn, we have: - fit(X, y, sample_weight) - predict(X) - score(X) Those are probably the main API we want to "override". In causal...

Along the lines of causal trees, adding some notes to be aware of: I've finished perusing and understanding the code in EconML about GRF and the general GRF paper. At...

Besides the GRF https://arxiv.org/pdf/1510.04342.pdf is a good paper

Possibly relevant papers: - https://proceedings.mlr.press/v70/hartford17a/hartford17a.pdf - https://eml.berkeley.edu/~powell/npiv.pdf

Note the DML API is probably highly relevant. Xref: #52

Honesty can be enabled either at the Python, or Cython level. If at the Python level, then we need to expose a `cpdef` API to allow the Tree to call...

Alternatively, doing it in Python will probably be simpler and we can just add this easily to the forked version of scikit-learn: 1. get subsamples for splitting and leaves 2....

@sampan501 and @yuxinB to help tackle honesty in trees and then add causaltree model, which adds the two options of fitting: - propensity model - double sampling

Okay now that I thought about this a little bit more, I think this can rely on the existing TreeBuilder/Splitter Cython code, where `y` can be 2D array. Other changes...

Okay! I will turn on auto merge for the first time then. Exciting 😱