Just wondering what is required so the colour for missing values is shown in the legend?
Looking at example from the UseR! ggplot2 book, p94
p <- qplot(sleep_total, sleep_cycle, data=msleep, colour=vore)
p + scale_colour_hue(na.value = "Black")
p + scale_colour_hue("What does \nit eat?", na.value="Black", breaks=c("herbi", "carni", "omni", "insecti", NA), labels=c("plants", "meat", "both", "insects", "don't know"))
the data point for vore=NA is shown in the plot but NA is not listed in the legend.
Thanks
Workaround for the problem would be to replace NA
values in your data with same other character (for example, unknown
) and plot data.
So, made new variable vore2 that has vore
values as characters. Then replaced NA
values with the unknown
.
msleep$vore2<-as.character(msleep$vore)
msleep$vore2[is.na(msleep$vore2)]<-"unknown"
In the plot used new variable vore2
for the colors.
p <- qplot(sleep_total, sleep_cycle, data=msleep, colour=vore2)
p + scale_colour_hue("What does \nit eat?",
breaks=c("herbi", "carni", "omni", "insecti", "unknown"),
labels=c("plants", "meat", "both", "insects", "don't know"))