search

judithabk6 / med_bench

BSD 3-Clause "New" or "Revised" License
8 stars 3 forks source link

Data simulation - Absence of confounder #3

Closed sami6mz closed 7 months ago

sami6mz commented 1 year ago

In get_simulated_data.py, simulate_data() :

If one sets dim_x = 0, no warging is returned. The output x is zero-dimensional, which does not prevent the other outputs from taking consistent values. I don't know whether this models the absence of a confounding variable, or whether it constitutes an error.

judithabk6 commented 7 months ago

Not sure we want that. Warning should be on the estimator side