Creating clustered bar chart with ggplot

[tronmed]

I would like to make a clustered/grouped bar chart with ggplot in R. However, I do not know how to make such a bar chart in the way I prefer. On the x-axis, I want to place 'drugset' and on the y-axis 'percentage'.

Here is an example of the data from which I want to make a clustered bar chart:

    Drugset     Percentage users
    Subset 1    58.63%
    Subset 2    76.25%
    Subset 3    27.27%
    Subset 4    30.43%

I would like to make one clustered bar chart where subset 1 and 2 are clustered (standing next to each other) and subset 3 and 4 as well. I am also wondering how to the change the colours.

Thanks for the help!

[nedallo]

below we can define our data frame :
df <- data.frame(Drugset=c("Subset1","Subset2","Subset3","Subset4"),percent=c(58.63,76.25,27.27,30.43),cat=c("cat1","cat1","cat2","cat2"))

df %>% ggplot(aes(Drugset,percent,fill=percent))+geom_bar(stat = "identity")+facet_grid(~cat)

see below image :

hope this helpful

[nedallo]

or you can have it in this way if you want :

df %>% ggplot(aes(Drugset,percent,fill=cat))+geom_bar(stat = "identity")

by giving two different colour

[tronmed]

This is very helpful, thankyou!
Is there also a way to get the two bars of subset 1 and 2 (with the pink/orange colour) more grouped together without the space in between?

[nedallo]

is this what you are looking for :

df %>% ggplot(aes(cat,percent,fill=Drugset))+geom_bar(stat = "identity",position = position_dodge())

Kind Regards

[tronmed]

Yes, this is exactly where I am looking for!
Lots of thanks!

Do you also know how to change the colours of the bars?

[FactOREO]

use the scale_fill_* functions from ggplot2 to modify the color of filled bar charts

[M_AcostaCH]

Hi @tronmed,

Im take the code maked for the other user and try to help you with the colors. You could use scale_fill_manual(). Like as they told you.

For select the best color you could use RGB tags. Select in this link.

Paletton - The Color Scheme Designer

For example, Im select four color, because the fill is about cat variable: 29BA12 , E1B715, 037590, and FF9D3C

Next you could add with the # in the script.

df %>% ggplot(aes(cat,percent,fill=Drugset))+
  geom_bar(stat = "identity",position = position_dodge())+
  scale_fill_manual(values=c("#29BA12","#E1B715","#037590","#FF9D3C")) +
  labs(title = 'your title') # for title

[tronmed]

Clear, thankyou! Is there a way to re-order the legenda? For example that you get: subset 3, subset 4, subset 1, subset 2?