How to create a stacked bar chart in r with ggplot

Question

My data is:



positive <- c("21", "22", "33", "21", "27") ##Percentage
negative<- c("71", "77", "67", "79", "73")  ##Precentage 
sample <- c("Hr", "Fi", "We", "Pa", "Ki")
mydata <- data.frame(positive , negative, sample)

I want to create a stacked bar graph that shows positive and negative percentages for each category in the sample variable. I tried this:

ggplot(mydata) +
  geom_bar(aes(x = sample, fill = positive))

but did not wrok. Sorry if the question looks basic. I started R a few weeks ago.

Does this answer your question? [Stacked bar chart](https://stackoverflow.com/questions/21236229/stacked-bar-chart) — zephryl, Mar 05 '22 at 05:16
You should be more specific, particularly what exactly didn't work, but also what you want this to look like. Your code is just going to show counts of rows by sample—is that what you want? — camille, Mar 09 '22 at 01:09

score 1 · Answer 1 · answered Mar 05 '22 at 05:28

You need to pivot your data to long format first (tidy format). Then, we can specify the x (sample), y (value), and fill (name) conditions.

library(tidyverse)

mydata %>%
  pivot_longer(-sample) %>%
  ggplot(aes(fill = name, y = value, x = sample)) +
  geom_bar(position = "stack", stat = "identity")

Output

score 1 · Accepted Answer · answered Mar 05 '22 at 05:28

This probably serves your purpose:

library(tidyverse)
mydata %>% pivot_longer(cols = !sample, names_to = "status", values_to = "percentage") %>% 
 ggplot(aes(fill = status, x = sample, y = percentage)) + 
 geom_bar(position = "stack", stat = "identity")

The result:

score 1 · Answer 3 · answered Mar 05 '22 at 06:03

Here is an alternative approach using values_transform to get numeric type and then using geom_col with position_fill and percentage features:

library(dplyr)
library(tidyr)
library(ggplot)
library(scales)

mydata %>% 
  pivot_longer(
    cols = -sample,
    names_to = "status",
    values_to = "percentage", 
    values_transform = list(percentage = as.integer)
  ) %>% 
  ggplot(aes(x = sample, y=percentage, fill=status))+
  geom_col(position = position_fill()) +
  scale_y_continuous(labels = scales::percent) +
  geom_text(aes(label = percentage),
            position = position_fill(vjust = .5))

Rfanatic · Answer 4 · 2022-03-06T19:41:39.557

Another solution ios the use of the side-by-side bars.

Sample data:

values <- c(21, 22, 33, 21, 27, -71, -77, -67, -79, -73) ##Percentage
sample <- factor(c("Hr", "Fi", "We", "Pa", "Ki"))
mydata <- data.frame(values, sample)

Sample code:

(ggplot(mydata, aes(x = sample, y = values)) +
  geom_bar(
    stat = "identity", position = position_stack(),
    color = "white", fill = "lightblue") +
  coord_flip())+
  labs(x="Sample",y="Percentage")+
  theme_bw()

Plot:

Sample code to show values:

(ggplot(mydata, aes(x = sample, y = values)) +
  geom_bar(
    stat = "identity", position = position_stack(),
    color = "white", fill = "lightblue"
  ) +
    geom_text(aes(label = values))+
  coord_flip())+
  labs(x="Sample",y="Percentage")+
  theme_bw()

Or if you want to color the positive and negative bars differently

(ggplot(mydata, aes(x = sample, y = values)) +
  geom_bar(
    stat = "identity", position = position_stack(),
    fill= ifelse(mydata$values < 0,"#ffcccb","lightblue")
  ) +
    geom_text(aes(label = values))+
  coord_flip())+
  labs(x="Sample",y="Percentage")+
  theme_bw()

How to create a stacked bar chart in r with ggplot

4 Answers4