Rose diagram for migration data

Question

I am trying to create a rose diagram showing average trajectory angle and distance for each subset of cells. I want the angle around the rose diagram to be the trajectory angle and the length of the bar in the diagram to be the total displacement.

Here is a test data set of the mean angle and displacement per group.

testsum<-data.frame(Group=c(1,2,3),
                angle=c(0.78,1.04,2.094),
                displacement=c(1.5,2,1))

When I try to plot this in a circular method, my chart looks very wrong.

p1<-ggplot(testsum, aes(x=angle,y=displacement))+
  coord_polar(theta="x",start=0)+
  geom_bar(stat="identity",aes(color=Group,fill=Group),width=.01)+    
  scale_x_continuous(breaks=seq(0,360,60))

It gives me this graph for output.

When based on what the data says, it should look more like this (drawing of intended output).

It seems to be placing the angles incorrectly? Any idea what I am doing wrong?

In the expected diagram group 3 appears twice. Is that wrong? — MKR, Feb 14 '18 at 20:50

score 2 · Answer 1 · answered Feb 14 '18 at 23:48

Although MLavoie "beat me" by 20 minutes, I think some readibility can be added by using NISTunits package:

library(ggplot2)
library(NISTunits)

testsum <- data.frame(
  Group = c(1, 2, 3),
  angle = c(0.78, 1.04, 2.094),
  displacement = c(1.5, 2, 1)
)

testsum$angle = NISTradianTOdeg(testsum$angle)

ggplot(testsum, aes(x = angle, y = displacement)) +
  coord_polar(theta = "x", start = NISTdegTOradian(-90), direction = 1) +
  geom_bar(stat = "identity",
           aes(color = Group, fill = Group),
           width = 1) +
  scale_x_continuous(breaks = seq(0, 360, 10), limits = c(0, 360))

Result:

To clip the bottom half check this answer.

score 1 · Answer 2 · answered Feb 14 '18 at 22:45

Maybe you can use windRose function from open air package

library(openair)
    testsum<-data.frame(Group=c(1,2,3),angle=c(0.78,1.04,2.094),displacement=c(1.5,2,1))
    testsum$angle=180*testsum$angle/pi
windRose(testsum,ws="displacement",wd="angle",breaks=c(0.5,1,1.5,2),paddle=F,key.header = "My Rose",angle=10, 
     statistic="prop.mean",key.footer = "Displacement")

score 1 · Accepted Answer · answered Feb 14 '18 at 23:19

Maybe you can try this:

testsum$angle_b=180*testsum$angle/pi
#
ggplot(testsum, aes(x=angle_b,y=displacement))+
    geom_bar(stat="identity",aes(color=Group,fill=Group),width=1) +    
    scale_x_continuous(breaks=seq(0,360,10), limits=c(0,360)) + coord_polar(direction=1)

Rose diagram for migration data

3 Answers3