How to use directlabels to label lines in ggplot2-CodePudding

library(ggplot2)
library(directlabels)
mydat <- structure(list(Name = c("Ana", "Josh", "Bart", "Ana", "Josh", 
"Bart"), color_line = c("purple", "purple", "orange", "purple", 
"purple", "orange"), x = c(0.864864864864865, 0.810810810810811, 
0.472972972972973, 0.851351351351351, 0.702702702702703, 0.648648648648649
), y = c(0.702702702702703, 0.675675675675676, 0.797297297297297, 
0.797297297297297, 0.72972972972973, 0.635135135135135), Class = c("A", 
"A", "A", "B", "B", "B")), class = c("data.table", "data.frame"
), row.names = c(NA, -6L))

mydat
   Name color_line         x         y Class
1:  Ana     purple 0.8648649 0.7027027     A
2: Josh     purple 0.8108108 0.6756757     A
3: Bart     orange 0.4729730 0.7972973     A
4:  Ana     purple 0.8513514 0.7972973     B
5: Josh     purple 0.7027027 0.7297297     B
6: Bart     orange 0.6486486 0.6351351     B

I have the above data set, and I plotted the results as follows:

g <- ggplot(mydat, aes(x = x, y = y, color = Class))   
       theme_classic()   
       geom_line(mapping = aes(group = Name), color = mydat$color_line)   
       geom_point()   
       scale_color_manual(values=c("springgreen4", "royalblue3"))
g

Now, I would like to add the Name of each individual to each line. Something that might look like this:

The closest thing that I found is the angled.boxes in library(directlabels). You can see how it looks

CodePudding user response：

One option to achieve your desired result would be to use the geomtextpath package which adds lots of options to add direct labels to lines and even allows for curved text. For your use case you could simply replace geom_line by geomtextpath::geom_textline to add your labels.

Note: Additionally I slightly adjusted your code to make use of the color aes for the lines and the fill aes to color the points.

library(ggplot2)
library(geomtextpath)

pal_color <- c("purple", "purple", "orange")
names(pal_color) <- c("Ana", "Josh", "Bart")

pal_fill <- c("springgreen4", "royalblue3")
names(pal_fill) <- c("A", "B")

base <- ggplot(mydat, aes(x = x, y = y))   
  scale_color_manual(values = pal_color)  
  scale_fill_manual(values= pal_fill)  
  theme_classic()  
  guides(color = "none")

base  
  geomtextpath::geom_textline(
    aes(group = Name, color = Name, label = Name), textcolour = "black")  
  geom_point(aes(fill = Class), shape = 21, stroke = 0, size = 2)

Or using the offset and gap arguments you could add the labels on top of the lines:

base  
  geomtextpath::geom_textline(
    aes(group = Name, color = Name, label = Name),
    offset = unit(5, "pt"), gap = FALSE, textcolour = "black")  
  geom_point(aes(fill = Class), shape = 21, stroke = 0, size = 2)

CodePudding user response：

It's not ideal, but I did this a long time ago using some math and manual adjustments

mydat %>% 
  group_by(Name) %>% 
  mutate(
    posx = mean(x)*1.01,
    posy = mean(y)*1.01,
    angle = -60*diff(range(y))/diff(range(x))
    ) %>% 
  ggplot(aes(x = x, y = y, color = Class))   theme_classic()  
  geom_line(mapping = aes(group = Name), color = mydat$color_line)  
  geom_point()   scale_color_manual(values=c("springgreen4", "royalblue3")) 
  geom_text(aes(x = posx, y = posy, group = Name, label = Name, angle = angle),
            size = 6, show.legend = FALSE, color = "black")