Clara Bicalho

Gathering @lilymedina's and my thoughts on this: 1. Currently get_types()$types returns both logical vector or integer vector depending on the query. So `get_estimands(M1, using = "posteriors", queries = "(Y[X=1]>Y[X=0])-(Y[X=1]Y[X=0]") /...

Good point. On similar note, `cluster_sampling_designer(icc = 0)` currently not allowed. Should we accommodate the DGP to draw from `rnorm()` when icc = 0?

This would apply to: * `binary_iv` arguments `a`, `b`, `d`, `a_Y`, `b_Y`, `d_Y` * `block_cluster_two_arm`, `two_arm_covariate`, `two_arm` arguments `ate` and `treatment_mean` * `two_by_two` arguments `mean_...` and `outcome_means`

Good point, this also applies to `factorial` argument `sd` and `outcomes_sd`. I think the existing checks/messages in `block_cluster_two_arm` handle well the more complex cases of `N` and `sd`.

Working on expanding Imbens (2014) eq 4.4 to multiple instruments and not sure if we want to assume complier types are constant or to calculate eq 4.4 late as diff-in-diff...

Error fixed in fb4f5d642e1cde60eaf878aaeb30aa7bf120da99