CODEWITHSUNDEEP
rggplot2geom-bar
Markus
Markus

Reputation: 169

Why is adding `position = "dodge"` to my `geom_bar` leading to values being displayed incorrectly?

I have a data frame:

df <- data.frame(human = c(1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5,1,2,3,4,5),
                 stage = c("A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4", "A1", "A2", "A3", "A4"),
                 class = c(0,1,0,0,0,1,0,1,1,1,0,1,0,0,0,1,0,1,1,1,0,1,0,0,0,1,0,1,1,1,0,1,0,0,0,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0)
)

And want bar chart for each stage on the x-axis:

ggplot(df, aes(x = stage, y = class, fill = as.factor(human))) + geom_bar(stat = "identity") + scale_y_continuous(limits = c(0,15))

enter image description here

Looks good, but I would like the human elements side by side, so I add position = "dodge":

ggplot(df, aes(x = stage, y = class, fill = as.factor(human))) + geom_bar(stat = "identity", position= "dodge") + scale_y_continuous(limits = c(0,15))

While colums are side by side now, for some reason all class = 1: enter image description here

Upvotes: 0

Views: 325

Answers (3)

MSR
MSR

Reputation: 2901

A solution that avoids the dplyr preprocessing to use stat_summary:

ggplot(df, aes(x = stage, 
               y = class, 
               fill = as.factor(human))) + 
  stat_summary(geom = "bar", 
               position = "dodge", 
               fun.y = "sum")

Upvotes: 1

LaLa
LaLa

Reputation: 135

It's because your "identies" are 0 or 1. One way to deal with this is to summarize your data before you plot it. For example:

library(tidyverse)

df %>% 
    group_by(human, stage) %>% 
    summarise(class = sum(class)) %>% 
    ggplot(aes(x = stage, y = class, fill = as.factor(human))) + 
    geom_bar(stat = "identity", position= "dodge")

enter image description here

Upvotes: 4

erocoar
erocoar

Reputation: 5923

Because you use stat = "identity". So then you'll have to count beforehand.

library(tidyverse)
df %>%
  count(stage, class, human) %>%
  ggplot(aes(x = stage, y = n, fill = as.factor(human))) + 
  geom_bar(stat = "identity", position = "dodge")

Upvotes: 0

Related Questions