0

I have data that is mostly centered in a small range but there is a significant number of points which are out of this range. I would like to plot a histogram for this data that will focus on the small range but will also show the outliers. Something like a log-scale for the histogram compare to this post

require(latticeExtra)
n<-1000
m<-matrix(c(rnorm(n,240,56),rnorm(n,47,13),rnorm(n,23,7),rnorm(n,8.7,3)),nrow=n)
m<-rbind(m, matrix(c(rnorm(n,385,78), rnorm(n,160,32), rnorm(n,4,.8), rnorm(n,.6,.12)), nrow=n))
grid<-data.frame(ev=m,model=rep(c("A","B"),each=n))

  histogram(~c(ev.1,ev.2,ev.3,ev.4)|model
      ,data=grid
      ,layout=c(2,1), xlab = "",ylab=""
      ,type=c("density")
      #,par.settings = list(superpose.polygon=list(alpha=c(.5,.5,.5,.5)))
      ,panel=function(x,subscripts=subscripts,...){
         panel.grid()
         x<-matrix(x,nrow=n)
         panel.histogram(x[,1],...,col=4,alpha=.5)
         panel.histogram(x[,2],...,col=2,alpha=.5)
         panel.histogram(x[,3],...,col=5,alpha=.5)
         panel.histogram(x[,4],...,col=7,alpha=.5)
         }

)
Community
  • 1
  • 1
Klaus
  • 1,946
  • 3
  • 19
  • 34
  • You've got a 2000 x 4 matrix here. Are you saying that it's meant to be a sample of size 8000 from a single population? – Hong Ooi Jul 07 '13 at 11:04
  • 2
    Could you please explain what you need that can't be achieved with the solution proposed in the post that you refer to? – QuantIbex Jul 07 '13 at 12:00
  • Each histogram in the plot has the same problem like the linked post. – Klaus Jul 07 '13 at 16:16

0 Answers0