2

I can't seem to figure out why this code isn't working for me. I am trying to add different geom_text labels to each facet of this graph. I've created a df for the annotation information.

Here is a sample of my dataset and code:

testParv <- data.frame(Stock = c("Birk","Dolly","Chil", "Pitt", "Birk","Dolly","Chil", "Pitt", "Birk","Dolly","Chil", "Pitt"), Parvicapsula = c("0", "1", "0", "1", "1", "0","0", "1", "0", "1", "1", "0"), Weight = c("19.2", "10.1", "5.2", "10.9", "20.5", "4.3", "15", "24", "6", "3", "8", "4.5"))

annotest <- data.frame(x1 = c(35, 35, 35, 35), y1 = c(0.4, 0.4, 0.4, 0.4), Stock = c("Birk", "Chil", "Dolly", "Pitt"), lab = c("p = 0.968", "p = 0.384", "p = 0.057", "p = 0.005"))

ggplot(data = testParv, aes(x = Weight, linetype = Parvicapsula)) +
  stat_ecdf() + 
  geom_text(data = annotest, aes(x = x1, y = y1, label = lab)) +
  theme_cowplot() + 
  facet_wrap(~Stock) + 
  ylab("ECDF")

This produces an error: Error in FUN(X[[i]], ...) : object 'Parvicapsula' not found, but when I take out the geom_text() code it works fine.

Does anyone know why this doesn't work?

Many thanks, Julia

JuliaB
  • 23
  • 2
  • Try leaving the first `ggplot()` call empty and moving all those parameters inside the `stat_ecdf()` call. `ggplot() + stat_ecdf(data = testParv, aes(x = Weight, linetype = Parvicapsula)) + ...` – nniloc Aug 27 '20 at 16:44
  • 1
    Thank you @nniloc, that does work! Good to know! – JuliaB Aug 27 '20 at 16:52

2 Answers2

0

I would suggest next approach. I made slight changes to label coordinates. The issue is that if you use geom_text() with other dataframe, you have to be careful that all variables from the dataframe in initial aes are also in the label dataframe. Here the code:

library(ggplot2)
#Data
testParv <- data.frame(Stock = c("Birk","Dolly","Chil", "Pitt", "Birk","Dolly",
                                 "Chil", "Pitt", "Birk","Dolly","Chil", "Pitt"),
                       Parvicapsula = c("0", "1", "0", "1", "1", "0","0", "1", "0", "1", "1", "0"),
                       Weight = c("19.2", "10.1", "5.2", "10.9", "20.5", "4.3",
                                  "15", "24", "6", "3", "8", "4.5")
                       )
#Labels
annotest <- data.frame(Weight = c(6, 6, 6, 6),
                       y1 = c(0.4, 0.4, 0.4, 0.4),
                       Stock = c("Birk", "Chil", "Dolly", "Pitt"),
                       lab = c("p = 0.968", "p = 0.384", "p = 0.057", "p = 0.005"),
                       Parvicapsula=NA)
#Plot
ggplot(data = testParv, aes(x = Weight, linetype = Parvicapsula)) +
  stat_ecdf() + 
  geom_text(data = annotest, aes(x = Weight, y = y1, label = lab)) +
  facet_wrap(~Stock) + 
  ylab("ECDF")

Output:

enter image description here

Duck
  • 39,058
  • 13
  • 42
  • 84
0

When you set the aes inside the ggplot() call those aesthetics will be used across all subsequent geoms. For example

ggplot(mtcars, aes(hp, mpg)) + geom_point() + geom_line()

In your case linetype from the first aes is being carried into geom_text. Several different ways to fix the error. Here are two options:

# Move the data and aes calls into stat_ecdf 
ggplot() +
  stat_ecdf(data = testParv, aes(x = Weight, linetype = Parvicapsula)) + 
  geom_text(data = annotest, aes(x = x1, y = y1, label = lab)) +
  theme_cowplot() + 
  facet_wrap(~Stock) + 
  ylab("ECDF") 

# In geom_text overwrite the linetype with NA
ggplot(data = testParv, aes(x = Weight, linetype = Parvicapsula)) +
  stat_ecdf() + 
  geom_text(data = annotest, aes(x = x1, y = y1, label = lab, linetype = NA)) +
  theme_cowplot() + 
  facet_wrap(~Stock) + 
  ylab("ECDF")
nniloc
  • 4,128
  • 2
  • 11
  • 22