# [R] Confidence Interval for p1-p2 and plot the CI with bar chart

Jim Lemon drj|m|emon @end|ng |rom gm@||@com
Sun Nov 14 02:47:37 CET 2021

```Hi Abou,
Perhaps this will be helpful. Be aware that you will cop some flak for
putting error bars on a bar plot.

v1=sample(0:1,100,TRUE),
v2=sample(0:1,100,TRUE),
v3=sample(0:1,100,TRUE),
v4=sample(0:1,100,TRUE),
v5=sample(0:1,100,TRUE))
ggps<-function(x,group) {
gns<-as.vector(table(group))
return(by(x,group,sum)/gns)
}
testggps<-data.frame(
group=c("A","A","A","B","B","B","B","C","C","C","C","C"),
x=c(1,0,1,1,0,1,0,1,1,0,0,0))
library(plotrix)
barpos<-barp(aaprop,ylim=c(0,0.65),col=c(2,3),names.arg=colnames(aaprop))
legend(2.5,0.65,c("Con","Exp"),fill=c(2,3))
dispersion(barpos\$x,barpos\$y,ulim=aaprop/10)

Jim

On Sun, Nov 14, 2021 at 11:01 AM AbouEl-Makarim Aboueissa
<abouelmakarim1962 using gmail.com> wrote:
>
> Dear All:
>
>
>
> I do have a binary data set with multiple variables, event = 1 in all
> variables. As an example, I attached a data set with 6 variables. The first
> column is the grouping variable. Then the next 5 columns are the binary
> data for 5 variables.
>
>
>
> - Can we compute the confidence interval for the difference between the two
> proportions of the event = 1 in both groups (say: G1 – G2) for the 5
> variables in one shut.
>
>
>
> - I also need to create the Bar plot of individual proportions (both groups
> side-by-side) and add the confidence intervals bar for the 5 variables in
> one graph.
>
>
>
>
>
> sep=",")
>
> Example.Data
>
> attach(Example.Data)
>
>
>
>
>
>
>
> #### For example, this is how I use the prop.test() function to get the CI
> for p1-p2
>
>
>
> x12 <- c(x1, x2)
>
> n12 <- c(n1, n2)
>
> prop.test(x12, n12, conf.level = 0.95)\$conf.int
>
>
>
> But, I am not sure how to use it for raw data, and for multiple pairs of
> data in one shut if possible.
>
>
>
> With many thanks in advance
>
> Abou
> ______________________
>
>
> *AbouEl-Makarim Aboueissa, PhD*
>
> *Professor, Statistics and Data Science*