Given frequencies in bins of given width, the comparable
frequency density is, I think,
sample size * probability density * bin width
So, you should specify bin width when doing the histogram
and superimpose an equivalent curve generated by -kdensity-
or one of its Ben Jann clones.
Checking especially welcome.
Nick
[email protected]
Daniel Schneider
> I have been trying to combine a kdensity graph and a histogram of the
> same distribution, but I would like to have the axis for the histogram
> in frequencies, not densities. I have experimented with using
> individual
> axes for both the histogram and the density line, but then the two
> results don't match anymore.
>
> Here is a command I have been using:
>
> _kdens variable ,kernel(triangle) ll(0) gen(y x)
> twoway (hist variable ,freq ) (line x y ,lcolor(black) lstyle(solid)
> yaxis(2) ytitle("",axis(2)) ylabel(none,axis(2)) ) , legend(off)
>
>
> Is there any solution for this problem? Using density and
> "faking" (i.e.
> relabeling) the axis for the histogram?
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
