Split barplot by grouping by days - r

Question

I have the following bar chart produced using this code:
MD1<-read.csv("MD_qual_OTU_sorted.csv")
MD1<-data.frame(Samples=c("A","B","C","D","E","F","G","H","I","J","K","L","M", "N","O","P","Q", "R"), Number.of.OTUs=c(13,10,9,9,15,11,7,7,9,9,5,10,10,7,15,17,8,9))
par(las=1)
barplot(MD1[,2],names.arg=MD1[,1], ylab='OTU Count', yaxt='n', xlab='MD samples', main='Total OTU count/Sample',density=c(90,90, 90, 90, 90, 90, 10, 10, 10, 10, 10, 10, 40, 40, 40, 40, 40, 40), col=c("yellow","yellow","pink", "pink","green","green","red","red", "purple", "purple", "blue", "blue", "orange", "orange","cyan", "cyan","chartreuse4", "chartreuse4" ))
usr <- par("usr")
par(usr=c(usr[1:2], 0, 20))
axis(2, at=seq(0,20,5))
I want to split samples A-F into a separate group (Day 3), G-L (Day 5) and M-R (Day 15)
There are similar questions posted however I am not sure how to tidy up the manner in which I have inputted my data to be able to use these solutions.

Answer 1

You could consider using ggplot2, separate plots are very easy using facet_wrap and facet_grid.
library(ggplot2)
#create a grouping variable
MD1$Day <- rep(c("Day 03","Day 05","Day 15"),
each=6)
p1 <- ggplot(MD1, aes(x=Samples,y=Number.of.OTUs)) +
geom_bar(stat="identity") + facet_wrap(~Day,
scales="free_x")
p1
Or, if you want to use base-R and approach your original image:
#add colors/densities
MD1$col <- c("yellow","yellow","pink", "pink","green","green","red","red",
"purple", "purple", "blue", "blue", "orange", "orange","cyan", "cyan","chartreuse4", "chartreuse4" )
MD1$density <- c(90,90, 90, 90, 90, 90, 10, 10, 10, 10, 10, 10, 40, 40, 40, 40, 40, 40)
#set 1 row three cols for plotting
par(mfrow=c(1,3))
#split and plot
lapply(split(MD1, MD1$Day),function(x){
barplot(x[,2],
names.arg=x[,1],
ylab='OTU Count',
ylim=c(0,20),
main=unique(x$Day),
col=x$col,
density=x$density)
})