r ggplot x axis tick marks order

3
votes

I've seen several questions about the order of x axis marks but still none of them could solve my problem. I'm trying to do a density plot which shows the distribution of people by percentile within each score given like this

library(dplyr); library(ggplot2); library(ggtheme)
ggplot(KA,aes(x=percentile,group=kscore,color=kscore))+
  xlab('Percentil')+ ylab('Frecuencia')+ theme_tufte()+ ggtitle("Prospectos")+
  scale_color_brewer(palette = "Greens")+geom_density(size=3)

but the x axis mark gets ordered like 1,10,100,11,12,..,2,20,21,..,99 instead of just 1,2,3,..,100 which is my desired output enter image description here

I fear this affects the whole plot not just the labels

1
rplotggplot2
Your x variable is a factor. You probably want it to be numeric. KA$percentile = as.numeric(as.character(KA$percentile)). – Gregor Thomas
adding a dput(head(KA)) would help confirm this – jeremycg
...but when your sorting is clearly alphabetical "1, 10, 100, 11, 12, ..., 2", confirmation is hardly necessary. – Gregor Thomas
And with 100 levels dput(droplevels(head(KA))) would be better. – Gregor Thomas
Because the x axis is ordered alphabetically you can check this post: stackoverflow.com/questions/12774210/… – mpalanco

1 Answers

1
votes

I'll turn my comment to an answer so this can be marked resolved:

Your x variable is (almost certainly) a factor. You probably want it to be numeric.

KA$percentile = as.numeric(as.character(KA$percentile))

When you're seeing weird stuff, it's good to check on your data. Running str(KA) is a good way to see what's there. If you just want to see classes, sapply(KA, class) is a nice summary.

And it's a common R quirk that if you're converting from factor to numeric, go by way of character or you risk ending up with just the level numbers:

year_fac = factor(1998:2002)
as.numeric(year_fac) # not so good
# [1] 1 2 3 4 5
as.numeric(as.character(year_fac)) # what you want
# [1] 1998 1999 2000 2001 2002