Add circular axis on circular plot - r

Question

I would like to overlay my plot with circles as axis to illustrate probability levels (e.g 0.25; 0.75, 1).
To reproduce the graphic you need these 2 csv files in working directory
https://drive.google.com/open?id=1RsleBYQFlm3ce3xuqTLK-_r9s374yd40
Or since I have been kindly advised by #Gregor in comments here are the headers of my data objects so no downloading is necessary:
dput(head(data))
structure(list(id = 1:6, individual = structure(c(1L, 12L, 23L,
26L, 27L, 28L), .Label = c("Person 1", "Person 10", "Person 11",
"Person 12", "Person 13", "Person 14", "Person 15", "Person 16",
"Person 17", "Person 18", "Person 19", "Person 2", "Person 20",
"Person 21", "Person 22", "Person 23", "Person 24", "Person 25",
"Person 26", "Person 27", "Person 28", "Person 29", "Person 3",
"Person 30", "Person 31", "Person 4", "Person 5", "Person 6",
"Person 7", "Person 8", "Person 9"), class = "factor"), value = c(0.658333333,
0.958333333, 0.720833334, 0.883333333, 0.779166667, 0.9375),
group = structure(c(1L, 1L, 1L, 1L, 1L, 1L), .Label = "A", class = "factor")), .Names = c("id",
"individual", "value", "group"), row.names = c(NA, 6L), class = "data.frame")
and second object:
dput(head(label_data))
structure(list(id = 1:6, individual = structure(c(1L, 12L, 23L,
26L, 27L, 28L), .Label = c("Person 1", "Person 10", "Person 11",
"Person 12", "Person 13", "Person 14", "Person 15", "Person 16",
"Person 17", "Person 18", "Person 19", "Person 2", "Person 20",
"Person 21", "Person 22", "Person 23", "Person 24", "Person 25",
"Person 26", "Person 27", "Person 28", "Person 29", "Person 3",
"Person 30", "Person 31", "Person 4", "Person 5", "Person 6",
"Person 7", "Person 8", "Person 9"), class = "factor"), value = c(0.658333333,
0.958333333, 0.720833334, 0.883333333, 0.779166667, 0.9375),
group = structure(c(1L, 1L, 1L, 1L, 1L, 1L), .Label = "A", class = "factor"),
hjust = c(0, 0, 0, 0, 0, 0), angle = c(84.375, 73.125, 61.875,
50.625, 39.375, 28.125)), .Names = c("id", "individual",
"value", "group", "hjust", "angle"), row.names = c(NA, 6L), class = "data.frame")
And then run following:
library(tidyverse)
library(ggplot2)
library(plotrix)
data=read.csv(file="data_object_2.csv", header=TRUE, sep=",")
label_data=read.csv(file="label_data_object_2.csv", header=TRUE, sep=",")
empty_bar=1
to_add = data.frame( matrix(NA, empty_bar*nlevels(data$group), ncol(data)) )
colnames(to_add) = colnames(data)
to_add$group=rep(levels(data$group), each=empty_bar)
data=rbind(data, to_add)
data=data %>% arrange(group)
data$id=seq(1, nrow(data))
number_of_bar=nrow(label_data)
angle= 90 - 360 * (label_data$id-0.5) /32
label_data$hjust<-ifelse( angle < -90, 1, 0)
label_data$angle<-ifelse(angle < -90, angle+180, angle)
p = ggplot(data, aes(x=as.factor(id), y=value)) +
geom_bar(stat="identity", fill=alpha("skyblue", 0.7)) +
ylim(-0.3,1) +
theme_minimal() +
theme(
axis.text = element_blank(),
axis.title = element_blank(),
panel.grid = element_blank(),
plot.margin = unit(rep(-1,4), "cm")
) +
coord_polar(start = 0) +
geom_text(data=label_data, aes(x=id, y=value, label=individual, hjust=hjust), color="black", fontface="bold",alpha=0.6, size=2.5, angle= label_data$angle, inherit.aes = FALSE ) +
geom_vline(xintercept = 0, color = "grey", linetype = "dashed") +
annotate("text", label = "p=0", x = 0, y = 0, color = "black") +
annotate("text", label = "p=1", x = 0, y = 1, color = "black") +
annotate("text", label = "p=0.5", x = 0, y = 0.5, color = "black")
p
It will result in this:
https://drive.google.com/open?id=1xDOym_nn-x9nrUoKpB9rtg7h7NYIfucF
I would like to overlay with circles indicating probability levels to enhance readability. All the on-line help which I have found is related to common Cartesian graph or geom_circle function which did not work either.
I will really appreciate any help.
Thanks Marek

Answer 1

Here's an option over a simplified version of your code (I didn't want to recreate all the labels, etc, just a pared down version of your chart). I thought about it like this: if this were in regular Cartesian coordinates, you could show an overlay of a probability by making a horizontal line, so in polar coordinates, that line would become a circle. Adding a geom_hline gives you a circle at whatever yintercept you set.
It might be good to label those probabilities; you can figure out what's the best way to do that in your context, but I just made a couple of circles, set the y-breaks to those same values, and moved the y axis title to be near the labels so they had a little explanation. Based on your context, that might not all be necessary.
As an aside, I'd recommend combining these two data frames into one, so you can more easily keep track of things and not have to set different data = arguments in different geoms.
library(tidyverse)
label_data %>%
ggplot(aes(x = individual, y = value)) +
geom_col(width = 0.5, fill = "skyblue", alpha = 0.7) +
geom_hline(yintercept = c(0.5, 0.75, 0.9), color = "gray60") +
scale_y_continuous(limits = c(-0.3, NA), breaks = c(0.5, 0.75, 0.9)) +
theme_minimal() +
theme(panel.grid = element_blank(), axis.title.y = element_text(hjust = 0.87)) +
coord_polar(start = 0) +
labs(x = NULL, y = "Probability")
Created on 2018-06-03 by the reprex package (v0.2.0).