ggplot stacked bar graph with 2 columns

Question

I'd like to create a stacked bar graph of my data frame df without having to transform the data, for interpretability reasons. My data looks like this:

#Code
year <- c(1:5)
burglaries <- c(234,211,201,150,155)
robberies <- c(12, 19,18,23,25)
total <- burglaries + robberies
df <- data.frame(year, burglaries, robberies, total)

#Output
print(df)

  year burglaries robberies total
1    1        234        12   246
2    2        211        19   230
3    3        201        18   219
4    4        150        23   173
5    5        155        25   180

I can create the plot I need by transforming my data set as follows:

df2 <- rbind(
        data.frame(year, "count" = burglaries, "type"="burglaries"),
        data.frame(year, "count" = robberies, "type"="robberies")
)

ggplot(df2, aes(x=year, y=count, fill=type)) +
    geom_bar(stat="identity")

Is there a way to create the same plot with the data frame df? Although I could transform the data, I worry it'll make it harder to follow what's happening in the program and catch errors (the data set I'm using is quite large).

You're working against the way ggplot was specifically designed to be used. ggplot wants you to transform (melt, gather, tidy, whatever you want to call it) your data first. So the short answer is no, not really. — joran
– joran, Commented Dec 1, 2016 at 17:21

heyydrien · Accepted Answer · 2016-12-01 21:20:00Z

3

I did some additional research and found the plot_ly() function from the library plotly allows you to do just that. Here's the link for more info: plotly website

plot_ly(data=df, x = ~year, y = ~burglaries, type = 'bar', name = 'Burglaries') %>%
    add_trace(y = ~robberies, name = 'Robberies') %>%
    layout(yaxis = list(title = 'Count'), barmode = 'stack')

answered Dec 1, 2016 at 21:20

heyydrien

9911 gold badge12 silver badges30 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Jacob · Accepted Answer · 2016-12-01 17:23:21Z

2

Ultimately a transformation is required but a more elegant way is to use tidyr:

df %>% 
   select(-total) %>% 
   gather(type, count, burglaries:robberies) %>% 
   ggplot(., aes(x=year, y=count, fill=forcats::fct_rev(type))) +
   geom_bar(stat="identity")

answered Dec 1, 2016 at 17:23

Jacob

3,6273 gold badges21 silver badges32 bronze badges

Collectives™ on Stack Overflow

ggplot stacked bar graph with 2 columns

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related