R: How to add normal distributions to overlapping grouped histograms with lattice

Question

I've been searching for ways to make overlapping grouped histograms with the function 'histogram' in lattice, which I've found an answer to here.

histogram( ~Sepal.Length,
 data = iris,
 type = "p",
 breaks = seq(4,8,by=0.2),
 ylim = c(0,30),
 groups = Species,
 panel = function(...)panel.superpose(...,panel.groups=panel.histogram,
                      col=c("cyan","magenta","yellow"),alpha=0.4),
 auto.key=list(columns=3,rectangles=FALSE,
               col=c("cyan","magenta","yellow3"))
 )

Now my question is if you could still add normal distributions for every group to this plot.

Possibly using this?

panel.mathdensity(dmath = dnorm, col = "black",
              args = list(mean=mean(x),sd=sd(x)))

end result should end up looking similar to this: image

score 0 · Answer 1 · answered Dec 04 '15 at 00:53

This is the closest I was able to get. The hint I used was here. My problem is that the density plot gets hidden behind the next histogram plot.

plot1 <- histogram( ~Sepal.Length,
                    data = iris,
                    type = "p",
                    ylim = c(0,30),
                    breaks = seq(4,8,by=0.2),
                    groups = Species,
                    col=c("cyan","magenta","yellow"),
                    panel = panel.superpose,
                    panel.groups = function(x,y, group.number,...){
                      specie <- levels(iris$Species)[group.number]         
                      if(specie %in% "setosa"){
                        panel.histogram(x,...)
                        panel.mathdensity(dmath=dnorm,args = list(mean=mean(x), sd=sd(x)), col="black")
                      }
                      if(specie %in% "versicolor"){
                        panel.histogram(x,...)
                        panel.mathdensity(dmath=dnorm,args = list(mean=mean(x), sd=sd(x)), col="black")
                      }
                      if(specie %in% "virginica"){
                        panel.histogram(x,...)
                        panel.mathdensity(dmath=dnorm,args = list(mean=mean(x), sd=sd(x)), col="black")
                      }

                    }       
                      )

R: How to add normal distributions to overlapping grouped histograms with lattice

1 Answers1