different colours in GGPLOT

[saurabh]

I have a dataset here.

data.frame(
ID = c(1,2,3,4,5,6,7,8,9,10,11,12,13,
14,15,16,17,18,19,20,21,22),
data.frame(
stringsAsFactors = FALSE,
ID = c(1,2,3,4,5,6,7,8,9,10,
11,12,13,14,15,16,17,18,19,20,21,22),
Rescan1 = c("No","No","No","No","No",
"No","No","No","No","Yes","No","No","No","No",
"No","No","No","No","No","No","No","Yes")

I'm trying to use ggplot to get two different colours for each bar chart that appears. I enter the following code
ggplot(prostate_cleaned, aes(Rescan1)) + geom_bar()

ID - Patient ID, Rescan1- whether or not patient needed rescanned (a discrete variable)

this gives me two gray bar chats (which I presume is the default). .
I want the 'Yes' to be in Green and 'No' to be in red.

I tried this:
ggplot(prostate_cleaned, aes(Rescan1)) + geom_bar() + scale_fill_manual(values = c("Red", "Green"))

but it doesnt change a think...Can someone help please?

[saurabh]

• Doesn't change a thing

[jrkrideau]

Can you repost that dataframe or provide it in dput() format? It makes no sense at the moment.

[saurabh]

Sure thing, I use datapasta:
data.frame(
stringsAsFactors = FALSE,
ID = c(1,2,3,4,5,6,7,8,9,10,
11,12,13,14,15,16,17,18,19,20,21,22),
Rescan1 = c("No","No","No","No","No",
"No","No","No","No","Yes","No","No","No","No",
"No","No","No","No","No","No","No","Yes")

[saurabh]

To further elucidate my problem...

ggplot(prostate_cleaned, aes(Rescan1)) + geom_bar(fill='green')

fills both graphs in green..but I want one in Red and the other in green. Thanks

[jrkrideau]

Maybe

dat1 <-   data.frame( ID = c(1,2,3,4,5,6,7,8,9,10,
         11,12,13,14,15,16,17,18,19,20,21,22),
  Rescan1 = c("No","No","No","No","No",
              "No","No","No","No","Yes","No","No","No","No",
              "No","No","No","No","No","No","No","Yes"))

ggplot(dat1, aes(Rescan1, fill = Rescan1)) + geom_bar()

BTW, that data.frame is still messed up bur easily fixed

[saurabh]

Thanks,

wondering if anyone has maanged to fix the problem?

[jrkrideau]

Where is the data in the datapasta coming from? All that is actually missing is the last ) in the data.frame

[saurabh]

data.frame(
stringsAsFactors = FALSE,
ID = c(1,2,3,4,5,6,7,8,9,10,
11,12,13,14,15,16,17,18,19,20,21,22),
Rescan = c(1,1,1,1,1,1,1,1,1,2,1,
1,1,1,1,1,1,1,1,1,1,2),
Rescan1 = c("No","No","No","No","No",
"No","No","No","No","Yes","No","No","No","No",
"No","No","No","No","No","No","No","Yes")
)

I've done and exact copy,paste from my PC using datapasta

[jrkrideau]

That looks fine now. Thanks

[nirgrahamuk]

in ggplot you map variable to aesthetics.
If you want to apply a fill, a fill aes needs to be provided, like this :

 ggplot(prostate_cleaned, aes(x=Rescan1,fill=Rescan1)) + 
  geom_bar() + 
  scale_fill_manual(values = c("Red", "Green"))

p.s. to format text as code in your forum posts, you use triple backticks to start and stop the formatting. like this

```
my code
```

code appears much easier to read when formatted this way.

[saurabh]

Many thanks, works...but when I want to display the reuslts in proportion format, I tend lose the colour. please see below

ggplot(prostate_cleaned, aes(x=Rescan1, fill=Rescan1, y = ..prop.., group = 1)) + geom_bar()+scale_fill_manual(values = c("green", "red")) + geom_text(stat= 'count', aes(label=..count.., vjust=1))

Also, not sure where to put the code in for reversibility (i.e. for the 'Yes' to appear first on the graph.

Thanks for all your patience..I've only been learning R for about a week. Problem solving helps me understand this programming more. I have no programming background.

[nirgrahamuk]

In general I recommend separating out calculations from plotting.
Here is a possible solution for what I believe you are trying to do

(prostate_cleaned_scaled <- 
  prostate_cleaned %>% group_by(Rescan1) %>%
  summarise(n=n()) %>% mutate(prop=n/sum(n)))

ggplot(prostate_cleaned_scaled,
       aes(x=Rescan1,
           fill=Rescan1, 
           y = prop)) + 
  geom_col()+
  scale_fill_manual(values = c("Green", "Red")) + 
  geom_text( aes(label=n,
                vjust=1))

[saurabh]

Many Thanks.

Worked. ..I've now learnt a bit about this 'piping' function thanks to your help.

I merely added:

scale_x_discrete(limits = c("Yes", "No")) 

to change the order of the barplots at the end the solution you provided

[mpsica]

I used this code

data.frame(stringsAsFactors = FALSE,
     ID = c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22),
     Rescan1 = c("No","No","No","No","No","No","No","No","No","Yes","No",
    "No","No","No","No","No","No","No","No","No","No","Yes")) %>% 
  ggplot(aes(Rescan1)) + 
     geom_bar(aes(fill=Rescan1))

... and the result is...

Tell me if that worked for you
N

[saurabh]

Thanks, I'd already found a solution as above.