Have you considered "truehist" in library(mass)?
Spencer Graves
p.s. This is discussed in Venables and Ripley (2002) Modern Applied
Statistics with S, 4th ed., which I recommend highly for anyone who uses
S-Plus or R. I only wish I had gotten the first edition when it
appeared.
Jim McLoughlin wrote:> Hi folks
>
> I'm trying to plot a normalized frequency histogram of some data.
After
> checking the docs, it seems there is no built in feature for this.
>
> from the definition for normalized frequency, I need to divide the
> relative frequency by the size of the intervals being used.
>
> So I could divide the series by this interval length, and then plot the
> relative frequency.
>
> The problem is determining the interval length that will be used in
> advance. I'm supposed to manually pick the # of intervals between 20
> and 40 (not use sturges). However, when I try different values for the
> nbreaks parameters, incrementing from 20 - 40, the histogram only
> changes at a couple of points on that interval.
>
> Anyone have any suggestions as to how I should create this plot? Or an
> explanation as to why my nbreaks parameter does not seem to force the
> number of intervals, but rather suggests a ballpark to be in?
>
> thanks
>
> Jim M
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help