Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Kernel Density Estimate Percentiles


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Kernel Density Estimate Percentiles
Date   Sun, 29 Jan 2012 14:21:18 +0000

I don't know of such a command and I think it's because it is not an
obvious route to follow.

-kdensity- smooths the empirical density function. To go from a
smoothed density function back to quantiles, you would need to
integrate the result to a distribution function and get its inverse.
But the results will inevitably differ slightly depending on choice of
kernel type and width.

-hdquantile- (SSC) offers one recipe for smoothing quantiles directly,
but based on different ideas.

That said, it's an unusual distribution if the 20, 50, 80% percentiles
differ much between original and smoothed distribution. This could
arise if the original  distribution took on just a few distinct
values, but in that case it is moot whether -kdensity- is a good idea.

Nick

On Sun, Jan 29, 2012 at 1:04 PM, Sakib bin Salam <sakibreed@gmail.com> wrote:

> I am interested in finding the percentiles values after a kernal
> density estimation.  I used Kdensity to generate a kernel density
> estimate. Is there a command that displays the 20th, 50th and 80th
> percentiles?
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index