r - How to properly add labels to ggplot2 horizontal boxplots

Question

Edited to provide reproducible example.

Sample data: https://owncloud.cesnet.cz/index.php/s/oopPE2Ut4quSVOK

Having this data

> head(df)
  Sample_Name        E Sentrix_ID Sentrix_ID_full
1    P129C1S1 5.636927 5058818037    5058818037_A
2    P129C1S1 5.794948 5058818037    5058818037_A
3    P129C1S1 5.608488 5058818037    5058818037_A
4    P129C1S1 5.989108 5058818037    5058818037_A
5    P129C1S1 5.570090 5058818037    5058818037_A
6    P129C1S1 5.555401 5058818037    5058818037_A

I am generating horizontal boxplots by

library(ggplot2)
library(ggthemes)

df <- read.csv("sample.csv")
df$Sentrix_ID <- as.factor(df$Sentrix_ID)
df$Sentrix_ID_full <- as.factor(df$Sentrix_ID_full)

head(df)

p <- ggplot(data = df,
            aes(y = E, x = Sentrix_ID_full, color = Sentrix_ID, label = Sample_Name)) +
  geom_boxplot(outlier.shape = NA) +
  scale_colour_manual(values = c("5058818037" = "red", "5226121006" = "green")) +
  theme_few() +
  guides(colour = FALSE) +
  labs(x = "Sentrix ID", y = "E", title = "intensity values") +
  coord_flip()

p

I want to add label to each boxplot:

p <- p + geom_text()

But this happens:

Is there any solution? Thank you in advance :)

With these kind of problems I mostly use the geom_text_repel() function from the ggrepel() package. Unfortunately I am not sure if this will also work with your problem. — Koot6133, Dec 01 '17 at 15:58
@Koot6133 Thanks, I will try it. But it seems each label is rendered multiple times (you can see some letters along boxplot). — jirinovo, Dec 01 '17 at 16:02
Please provide your dataset or a toy example with similar data features. See [here](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) for details. — Claus Wilke, Dec 01 '17 at 16:08
Please show the code that you used with `geom_text` (second plot) — pogibas, Dec 01 '17 at 16:10

score 5 · Accepted Answer · answered Dec 01 '17 at 16:34

One suggestion : use a second dataframe to label each boxplot only one time and at a defined position :

# mock data
set.seed(1)
df <- data.frame(var_1 = c(rep("A", 15), rep("B", 15), rep("C", 15)),
             value = sample(100, 45, replace= T))

# the kind of plot you have
ggplot(df, aes(x = var_1, y = value, label = var_1)) +
  geom_boxplot() +
  coord_flip() +
  geom_text()

# Use of plyr package
library(plyr)
df2 <- ddply(df, "var_1", summarize, max_value = max(value, na.rm = T))
# I chose max_value as position for the label

# the new plot
ggplot(df, aes(x = var_1, y = value)) +
  geom_boxplot() +
  coord_flip() +
  geom_text(data = df2, aes(x = var_1, y = max_value, label = var_1), hjust = -0.5) # use hjust to adjust horizontal position

Thanks, works fine :) I've just used third quartile (`quantile(E, probs = c(0.75), na.rm = T)`) instead of max. Label positions are adjusted with `nudge_y = 1` instead of `hjust`. Result: https://ctrlv.cz/shots/2017/12/04/Uka6.png — jirinovo, Dec 04 '17 at 16:39

r - How to properly add labels to ggplot2 horizontal boxplots

1 Answers1