***Getting Bootstrapped Standard errors for counts of binary outcomes. 
program myratio, rclass
	sum exp if study==1 & outcome==1 & cond==7
	local nom = r(N)
	sum exp if study==1 & cond==7
	local denom = r(N)
	return scalar ratio = `nom'/`denom'
end
bootstrap r(ratio), reps(1000): myratio

**Notes: There is probably a better way, but tabulate does not store the scalars needed and this works!
*First sum if statement denotes the outcome (yes or 1s) out of the sample population
*Second if statement denotes the sample population for the condition (cond) and the study (study)

****CREDIT Stata
*https://www.stata.com/features/overview/bootstrap-sampling-and-estimation/