Sorting Y Axis Values ggplot

Question

I'm trying to create a dotplot where countries are listed on my Y axis from A-Z top to bottom. The medal count will be the X axis for each of the four plots, one each for gold, silver, bronze, and total. Of course, ggplot prefers to plot countries from Z-A and despite reading all about the problem, I haven't resolved the issue. I appreciate any straightforward help on both the coding and comprehension fronts.

mdat <- melt(raw, value.name = "Count", variable.name = "Place", id.var = "Country")
mdat[, "Place"] <- factor(mdat[, "Place"], levels=c("Gold", "Silver", "Bronze",          "Total"))
##I know my problem is likely on or around the above line  ##
plot1 <- ggplot(mdat, aes(x = Count, y = Country, colour = Place)) +
  geom_point() +
  facet_grid(.~Place) + theme_bw()+
  scale_colour_manual(values=c("#FFCC33", "#999999", "#CC6600", "#000000")) 
print(plot1)


Algeria     Gold    4
Argentina   Gold    5
Armenia     Gold    1
Algeria     Silver  2
Argentina   Silver  5
Armenia     Silver  2
Algeria     Bronze  4
Argentina   Bronze  2
Armenia     Bronze  0

Welcome to SO. Please provide a [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) — agstudy, May 23 '14 at 06:19

Jaap · Accepted Answer · 2014-05-23T07:28:34.460

You have to sort the levels of Country before you plot. Also, there is no Total level the data you provided. The following appraoch should give you the desired result:

Reading the data (including a Total level for the Place variable):

mdat <- read.table(text="Country    Place Count
Algeria     Gold    4
Argentina   Gold    5
Armenia     Gold    1
Algeria     Silver  2
Argentina   Silver  5
Armenia     Silver  2
Algeria     Bronze  4
Argentina   Bronze  2
Armenia     Bronze  0
Algeria     Total   10
Argentina   Total   12
Armenia     Total   3", header=TRUE)

Sorting the levels of the Country variable:

mdat$Country <- factor(mdat$Country,levels=sort(unique(mdat$Country),decreasing=TRUE))

Getting your Place variable in the correct order:

levels(mdat$Place) <- c("Bronze"=3,"Gold"=1,"Silver"=2,"Total"=4)
mdat$Place <- as.numeric(mdat$Place)
mdat$Place <- as.factor(mdat$Place)
levels(mdat$Place) <- c("Gold","Silver","Bronze","Total")

Creating the plot:

ggplot(mdat, aes(x = Count, y = Country, colour = Place)) +
  geom_point(size=4) +
  facet_grid(.~Place) + theme_bw()+
  scale_colour_manual(values=c("#FFCC33","#999999","#CC6600","#000000"))

which gives the following plot: enter image description here

As you melted your data already, I suspect that there is no Total variable in the raw dataframe. You can calculte that with:

raw$Total <- rowSums(..specify the Gold, Silver & Bronze columns here..)

Thanks, Jaap. I actually did have a 'total' level and subsequent variable, but very good catch. Sorry, it was the end of the night. Your code worked perfectly. — SimplySnee, May 23 '14 at 14:56
By chance, does ggplot allow me to duplicate that beautifully sorted Y axis label on the right side? — SimplySnee, May 23 '14 at 14:56
The developer of `ggplot2` [doesn't like](http://stackoverflow.com/a/3101876/2204410) dual y-axis to put it mildly (and for good reasons). As a result there is no easy way to do this. The answers in [this question](http://stackoverflow.com/questions/15334494/how-to-change-positions-of-x-and-y-axis-in-ggplot2), or [this example](http://rpubs.com/kohske/dual_axis_in_ggplot2) seem to be the only way to get the desired result. — Jaap, May 23 '14 at 15:15
Interesting. Let's see if I learned anything. If I want to order the Y-axis by the value of "total", this seems to work, but am i right? mdat$Country <- factor(mdat$Country, levels=mdat[order(mdat$Count,decreasing=F),]$Country) — SimplySnee, May 24 '14 at 02:13

Sorting Y Axis Values ggplot

1 Answers1