No matter what setting I use for cut.mse in 'microsynth,' only 2 permutation groups are removed. I can see in the resulting graphs that several of the permutations have just awful matches in the pre-period, and yet they are still included in the final result. I've attached data to reproduce the issues I'm seeing, here's what I'm running
Hi,
No matter what setting I use for cut.mse in 'microsynth,' only 2 permutation groups are removed. I can see in the resulting graphs that several of the permutations have just awful matches in the pre-period, and yet they are still included in the final result. I've attached data to reproduce the issues I'm seeing, here's what I'm running
ms.nh6 <- microsynth(df, idvar="state_fips", timevar="quart_num", intvar="int", start.pre=start.pre, end.pre=end.pre, end.post=end.post2, result.var="rx_cap_alldrugs", match.out="rx_cap_alldrugs", check.feas=TRUE, use.backup = TRUE,confidence=0.95, cut.mse=0.01, perm=nrow(df)-1, jack=TRUE, test="two-sided", n.cores=FALSE, match.out.min=c("liver_restrict","sobriety_restrict","mco_rx"))
plot_microsynth(ms.nh6, all=NULL, start.pre=start.pre, end.pre=end.pre, end.post=end.post2, main.tc=main.tc, main.diff="Difference", xlab.tc="Quarter Time", xlab.diff="Quarter Time", ylab.tc="Rx per 100k enrollees", ylab.diff="Rx per 100k enrollees")
microsynth.zip