Counterfactual bayesian distribution regression

distreg draws randomly from the density of counterfactual of F(yo) at a threshold value yo

distreg_cfa(thresh, data0, MH = "IndepMH", cft, cfIND, ...)

Arguments

thresh	threshold value that is used to binarise the outcome variable
data0	original data set with the first column being the outcome variable
MH	Metropolis-Hastings algorithm to use; default:"IndepMH", alternative "RWMH"
cft	column vector of counterfactual treatment
cfIND	the column index(indices) of treatment variable(s) to replace with cft in data0
...	any additional inputs to pass to the MH algorithm

Value

robj a list of a vector of fitted values corresponding to random draws from F(yo), counterfactual F(yo), and the parameters

Examples

data0=faithful[,c(2,1)]; qnt<-quantile(data0[,1],0.25)
cfIND=2 #Note: the first column is the outcome variable. 
cft=0.95*data0[,cfIND] # a decrease by 5%
dist_cfa<- distreg_cfa(qnt,data0,cft,cfIND,MH="IndepMH",iter = 102, burn = 2)
#> IndepMH algorithm successful. Acceptance ratio =  0.4901961  
par(mfrow=c(1,2)); plot(density(dist_cfa$counterfactual,.1),main="Original")
plot(density(dist_cfa$counterfactual,.1),main="Counterfactual"); par(mfrow=c(1,1))