RE: st: RE: How to define shortest possible period with 95% of observations

["Nick Cox" <n.j.cox@durham.ac.uk>]

This isn't the question. For a start, -sumnmarize, detail- only reports certain percentiles. Even if it did report all possible percentiles (order statistics), the problem still remains: which of various intervals that contain at least 95% of the observations is the shortest? 

That is, which of the differences 

95% percentile - 0% percentile 

(95 + a bit)% - (0 + a bit)% percentile 

100% percentile - 5% percentile 

is the shortest? 

Nick 
n.j.cox@durham.ac.uk 

Dan Waldo

I should have said

 summarize date [fweight=fires] if year==`t' , detail

shouldn't I. Sorry.

--- On Tue, 5/11/10, Dan Waldo <dan_waldo@yahoo.com> wrote:

> I may have grievously oversimplified
> the issue raised, but is it possible that the summarize
> command would work? That is, to use the 5th and 95th
> percentiles from
>  
> .summarize date if fires>0 & year==1983, detail
> 
> This could be turned into a forvalues loop if there are
> many years:
> forval t=1978/1993 {
> summarize date if fires>0 , detail
> }
> 
> > Daniel Mueller
> >
> > I have a strongly unbalanced panel with 100,000
> observations
> > (=fire occurrences per day) that contain between none
> (no fire)
> > and 3,000 fires per day for 8 years. The fire events
> peak in
> > March and April with about 85-90% of the yearly
> total.
> >
> > My question is how I can define the shortest possible
> > continuous period of days for each year that contains
> 95%
> > of all yearly fires. The length  and width of the
> periods may
> > slightly differ across the years due to
> > climate and other parameters.
> >
> > I am sure there is a neat trick in Stata for this, yet
> I have not
> > spotted it. Any suggestions would be appreciated.

*
*   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/