LucyMcGowan
commented
1 year ago (for what it is worth I don't think this is going to solve our problems because it basically just gives weights very close to 0 so then the weighted model is very close to the unadjusted model 🤔 )
@malcolmbarrett can you take a peek at this? I added in the ATO weights (using an ordinal logistic model) but I can't get tar_make to run (and I can't figure out how to debug it) I think it is because sometimes the model doesn't converge? but I would have expected it to still run and just give predicted values close to 1 or 0 🤔