I have the following dataset
prg<-c('BRK1','BRK2','MAN1','QN')
comp<-c(5,0,12,9)
reg<-c(0,5,0,0)
term<-c(3,0,0,0)
df<-data.frame(prg,comp,reg,term)
df
prg comp reg term
1 BRK1 5 0 3
2 BRK2 0 5 0
3 MAN1 12 0 0
4 QN 9 0 0
Basically, I want to draw a barplot using ggplplot where prg is on the x axis and there are three bars for each of the prg in the form of comp reg and term representing the corresponding counts.
How can I go about doing this?
You have to melt it before plotting
library(tidyverse)
df.melted <- df %>% melt(id.var="prg")
# prg variable value
# 1 BRK1 comp 5
# 2 BRK2 comp 0
# 3 MAN1 comp 12
# 4 QN comp 9
# 5 BRK1 reg 0
# 6 BRK2 reg 5
# 7 MAN1 reg 0
# 8 QN reg 0
# 9 BRK1 term 3
# 10 BRK2 term 0
# 11 MAN1 term 0
# 12 QN term 0
## Then plot it.
## Use the `fill` parameter to get the breakdown
library(ggplot2)
ggplot(df.melted, aes(x = prg, fil=variable, y=value)) +
geom_bar(stat = "identity") +
ggtitle("StackOverflow Roolz!")
I agree with Ricardo Saporta, but I might add that without position = "dodge" the whole thing is difficult to understand.
library(tidyverse)
df.tidy <- gather(df, key, value, - prg)
ggplot(data = df.tidy, aes(x = factor(prg), y = value, col = key)) +
geom_bar(stat = "identity", position = "dodge")
this allows the bars to be spaced and distinguishable from each other.
Using reshape2 and ggplot2:
df %>%
melt(id.vars = "prg") %>%
ggplot(aes(prg, value)) +
geom_bar(aes(fill = variable), position = "dodge", stat = "identity")
