Add extra value to ggplot barplot

Question

I have the following data:

Sp  Type    Val1    Val2
A   One     200     50
A   Two     100     10
C   One     300     150
C   Two     200     10

I did the following to get stacked barplot:

ggplot() +
  geom_bar(data=test, aes(y = Val1, x = Sp, fill = Type), stat="identity",position='stack')

Hence, I get two stacked bars for A, B each with stacks of Type 1 & 2 (total size of A is 200+100 =300). Here, val2 is a fraction of unknowns in each Type. How do I overlay it in respective portions of stack? i.e in type A in Val1, fraction of unknown is Val2.

Thanks in Advance.

AP

Please be more precise. Do you want to overlay `Val2` above `Val1`. Or do you want a stacked barplot including both values? — Roman, Aug 02 '17 at 12:58
It is stacked barplot with Val2 shaded. I mean I want to add the Val2 inside Val1 bar. i.e A will have 2 stacks, 1 with 200, 2 with 100. In stack with 200, I want to shade a region with 50 and 100 with 10. It will be better, if we have 50 & 10 with same color so that, it would look like proportion marked in both stacks !. — Arun, Aug 03 '17 at 14:00
ok, then have a look at my answer! The others only summing up Val1 and Val2. — Roman, Aug 04 '17 at 06:00

Mal_a · Answer 1 · 2017-08-02T10:06:26.453

2

Is this what You are looking for?:

library(ggplot2)
data <- data.frame(Sp  = c("A","A","C","C"), 
                   Type = c("one","two","one","two"), 
                   Val1 = c(200,100,300,200),
                   Val2 = c(50,10,150,10))
library(reshape2)
data <- melt(data, id=c("Sp","Type"))
data$Type2 <- paste(data$Type, data$variable, sep="_")

[UPDATE] The data which you get after melting:

 Sp Type variable value    Type2
1  A  one     Val1   200 one_Val1 # it has value of 200 for Sp A
2  A  two     Val1   100 two_Val1 # it has value of 100 for Sp A
3  C  one     Val1   300 one_Val1
4  C  two     Val1   200 two_Val1
5  A  one     Val2    50 one_Val2
6  A  two     Val2    10 two_Val2
7  C  one     Val2   150 one_Val2
8  C  two     Val2    10 two_Val2

whereas one_Val1 is equal to 200 and two_Val1 to 100 --> 200 + 100 = 300

ggplot() +
  geom_bar(data=data, aes(y = value, x = Sp, fill = Type2), stat="identity",position='stack')

I have melted at first your data to get values of Val1 and Val2 in one column to use it further and paste it together with the Type column.

edited Aug 02 '17 at 10:06

answered Aug 02 '17 at 09:13

Mal_a

3,670
1
27
60

The total size of A is not summing up to 300 (200+100) with your solution. – Roman Aug 02 '17 at 09:59
one_Val1 and two_Val1 = 300 --> this is Sp: A for Type one – Mal_a Aug 02 '17 at 10:01
No thist is not true. – Roman Aug 02 '17 at 10:03
1

if it is wrong that i think you incorecctly summed the question, by saying that: you wanna overlay respective options of stack...it is actually not very clear – Mal_a Aug 02 '17 at 10:10

score 1 · Answer 2 · answered Aug 02 '17 at 09:22

1

In case you wanted it divided per Val1/Val2 values

library(ggplot2)
library(reshape2)
test <- melt(test, id=c("Sp","Type"))
ggplot(data=test) +
  geom_bar(aes(y = value, x = Sp, fill = Type), stat="identity",position='stack')+
  facet_wrap(~variable)

answered Aug 02 '17 at 09:22

Patrik_P

3,066
3
22
39

Roman · Accepted Answer · 2017-08-02T09:51:28.683

You can try:

d$xmin <- rep(c(0.55, 1.55),each=2)
d$xmax <- rep(c(1.45, 2.45),each=2)
d$ymin <- c(100, 0, 200, 0)
d$ymax <- c(150, 10, 350, 10)

ggplot(d) + 
    geom_col(aes(x=Sp, y=Val1, fill=Type)) +
    geom_rect(aes(xmin=xmin, xmax=xmax, ymin=ymin, ymax=ymax), alpha=0.5)

The idea is to manually add rectangles over the bars (I'm using here geom_col as this function uses stat_identity as default). Thus you calculate the mins and max by yourself and add some alpha to overplot the bars.

Or you can try a more automatic dplyr solution:

library(tidyverse)
d %>% 
  arrange(Sp, -as.numeric(Type)) %>% 
  mutate(ymin=ifelse(Type=="One",lag(Val1),0),
         ymax=ifelse(Type=="Two",Val2, lag(Val1)+Val2)) %>% 
  mutate(Sp_n=as.numeric(Sp)) %>% 
  ggplot() + 
  geom_col(aes(x=Sp_n, y=Val1, fill=Type))+
  geom_rect(aes(xmin=Sp_n-0.45, xmax=Sp_n+0.45, ymin=ymin, ymax=ymax),
  fill="white", alpha= 0.7) +
  scale_x_continuous(breaks = 1:2, labels = unique(d$Sp))

Add extra value to ggplot barplot

3 Answers3