Alter my values to surround a certain point R - r

Question

I have the following data, which shows the values for 5 different cohorts of patients (3 patients in each cohort):
dat <- data.frame(Cohort=c(1,1,1, 2,2,2, 3,3,3, 4,4,4, 5,5,5),
LEN_Dose=c(15,15,15, 25,25,25, 15,15,15, 10,10,10, 10,10,10),
DLT=c("N","N","N", "Y","Y","N", "Y","N","Y", "N","N","Y", "N","N","Y"))
I would like to modify the cohort levels to be +/- 0.2 of the main cohort number so they don't sit on top of one another in a graph. I can achive what I want like this:
dat$Cohort <- dat$Cohort-0.2
dat$Cohort <- ifelse(duplicated(dat$Cohort), dat$Cohort+0.2, dat$Cohort)
dat$Cohort <- ifelse(duplicated(dat$Cohort), dat$Cohort+0.2, dat$Cohort) # have to run this twice as there are 3 patients
So the result is:
head(dat)
# Cohort LEN_Dose DLT
# 0.8 15 N
# 1.0 15 N
# 1.2 15 N
# 1.8 25 Y
# 2.0 25 Y
# 2.2 25 N
But I'm wondering if there's a better way to do this? Eg somehow inputting the base cohort level and some function automatically works out the 3 values I need?
The point is to eventually graph the data using this graph:
ggplot(aes(x=Cohort, y=as.numeric(LEN_Dose)), data = dat) +
ylab("Dose Level\n") +
xlab("\nCohort") +
ggtitle("\n") +
scale_y_continuous(breaks = c(5, 10, 15, 25),
label = c("1.2mg/kg\n5mg", "1.2mg/kg\n10mg", "1.8mg/kg\n15mg", "1.8mg/kg\n25mg")) +
scale_fill_manual(values = c("white", "darkred"),
name="Had DLT") +
geom_line(colour="grey20", size=1) +
geom_point(shape=23, size=6, aes(fill=DLT), stroke=1.1, colour="grey20") + # 21 for circles
theme_classic() +
theme(legend.box.margin=margin(c(0,0,0,-10))) +
expand_limits(y=c(5,25))
EDIT: I have tried position = position_jitter, position = position_dodge and all the other types of positions within ggplot itself, but they don't space the points equally or in any particular order, which is why I'm trying to modify the dataframe itself

Answer 1

How about writing your jitter function, something like:
jitterit<- function(xTojitter= dat$Cohort, howMuchjitter=0.2){
x<-xTojitter
uni<-unique(x)
for (i in 1:length(uni)) {
if (is.na(uni[i])) {
x[is.na(x)]<-NA
} else if (sum(x==uni[i], na.rm = T) %%2 ==1) {
if(sum(x==uni[i], na.rm = T)==1){x[x==uni[i] & !is.na(x)][middle] <- uni[i]
} else {
middle<-ceiling (sum(x==uni[i], na.rm = T)/2)
x[x==uni[i] & !is.na(x)][1:(middle-1)] <- uni[i] - howMuchjitter
x[x==uni[i] & !is.na(x)][(middle+1):sum(x==uni[i], na.rm = T) ]<- uni[i] + howMuchjitter
x[x==uni[i] & !is.na(x)][middle] <- uni[i]
}} else if (sum(x==uni[i], na.rm = T) %%2 ==0) {
x[x==uni[i] & !is.na(x)]<- rep(c(uni[i] - howMuchjitter,uni[i] + howMuchjitter), each= sum(x==uni[i],na.rm = T)/2)
}
}
return(x)
}
It will work for all kind of duplicated data (even or odd number of duplication)
jitterit(xTojitter = c(1,1,2,1,2,NA), howMuchjitter=0.2)
[1] 0.8 1.0 1.8 1.2 2.2 NA