Search code examples
rggplot2geom-bargeom-col

Geom_errorbar in ggplot doesn't show the correct standard error

I want to add error bars to my geom_col ggplot. The error bar is indeed added to my plot, but the error bars are not correct. Can someone please help me figure out how to adapt my code? I think that the y-values is wrong. It should probably not be the length, but the actual values of the "Number"'s from my data frame. I apologize if this is a really simple question, i am quite new in R. Thank you in advance.

enter image description here

number_of_sprout <- data.frame(
  Condition = c("RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "RBP7 KO", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl", "Ctrl"), 
  Number = c(8, 11, 11, 17, 18, 13, 16, 4, 11, 9, 12, 11, 15, 11, 17, 8, 18, 9, 7, 11, 11, 12, 8, 12, 6, 8, 10, 19, 9, 14, 7, 9, 12, 11, 8, 12, 6, 8, 10, 8, 11, 11, 10, 7, 6, 6, 6, 9, 11, 10, 11, 12, 8, 14, 7, 8, 11, 9, 9, 12, 10, 7, 13, 17, 15, 9, 14, 11, 11, 8, 14, 13, 7, 8, 6, 7, 9, 7, 10, 7), 
  timepoint = c(24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48, 48)
)


sph1 <- ggplot(data = number_of_sprout, aes(x = factor(timepoint, ordered = TRUE), y=Number, fill = Condition)) +
  stat_summary(fun = mean, geom = "col", position = position_dodge()) +
  scale_fill_manual(values = c("slategray4", "thistle3")) +
  geom_point(position = position_dodge(width = 0.9)) +
    labs(x = "Hours", y = "Number of sprouts") +
    theme_classic() +
    ggtitle("Spheroids - Donor 37") +
    theme_bw() +
    theme(text=element_text(family="Times New Roman", face="bold")) +
    theme(aspect.ratio = 1.5/1)

se5 <- function(y) sd(y)/length(y)   # to calculate standard error in the mean
sph1+stat_summary(geom="errorbar",position=position_dodge(width=0.9),
                 fun.data=function(y)c(ymin=mean(y)-se5(y),ymax=mean(y)+se5(y)), width=0.2)
Solution

  • I assume you just have the wrong formula for the standard error of the mean. I have this saved in my personal utility package, so I don't need to look that up - cheekily stolen from this thread at some point.

    Of another note, perhaps check http://data-to-viz.com/caveat/error_bar.html for a more general caveat why error bars might be not ideal (and not very necessary if you plot the entire data, in general)

    # replace your function
se5 <- function(x, na.rm=FALSE) {
  if (na.rm) x <- na.omit(x)
  sqrt(var(x)/length(x))
}  # to calculate standard error in the mean
sph1+stat_summary(geom="errorbar",position=position_dodge(width=0.9),
                  fun.data=function(y)c(ymin=mean(y)-se5(y),ymax=mean(y)+se5(y)), width=0.2)