search

dpuelz / BicliqueRT

Randomization Tests of Causal Effects under General Interference
https://arxiv.org/pdf/1910.10862.pdf
6 stars 4 forks source link

Issue with the cluster interference example #17

Closed samzl1 closed 1 year ago

samzl1 commented 2 years ago

Hi,

Thanks for building this package! I have played around with it for a while and found an issue in the cluster interference example you have.

In the out_bassefeller function, the Z_obs should be an exposure mapping rather than the binary treatment vector since you are using (Zobs==2) to return the Yobs. However, in the cluster interference example, you are using Z in the out_bassefeller to construct Yobs, causing the potential outcomes for the spillover and control groups to be the same. So you might need to double-check on this part.

Best

dpuelz commented 2 years ago

Thanks for the feedback! @szhuang1, can you take a look into this?