Re: st: data management question

[Nick Cox <njcoxstata@gmail.com>]

You could -tsset- or -xtset- these data and then use -tsspell- (SSC)
to identify spells in which there was uninterrupted use.

tsset id day
tsspell, pcond(number)

Then the median length of spells is given by

su _seq if _end, detail

Principles for analysing spells are discussed in


SJ-7-2  dm0029  . . . . . . . . . . . . . . Speaking Stata: Identifying spells
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q2/07   SJ 7(2):249--265                                 (no commands)
        shows how to handle spells with complete control over
        spell specification

This paper is independent of -tsspell-. -tsspell- I suspect provides
all the basic tools you need.

Nick

On Wed, Feb 29, 2012 at 11:14 PM, Maria Cecilia Vieira da Silva
<mcvieira@mapigroup.com> wrote:
> Hello,
> My data set looks like below. Each patient (here patient 1 and 2) takes or not pills each day (from day 1 to day 12). I would like to estimate the median number of days that the patients take the pill uninterruptedly. For example, the median is 3 in the case below - Patient 1 (3, 4, 1) and Patient 2 (3, 2).
> Any suggestions about how to implement it in a large data set?
> I though about creating a dummy (number of pills>0) and performing sums for the series bordered by zeros, but I am not sure about a command that would allow me to create subseries defined as series of numbers bordered by zeros. Actually, if I could create such subseries, it would suffice to perform a count.
> Thanks,
> Maria Cecilia
>
>
> id day number of pills
> 1 1 1
> 1 2 2
> 1 3 1
> 1 4 0
> 1 5 0
> 1 6 2
> 1 7 1
> 1 8 2
> 1 9 3
> 1 10 0
> 1 11 0
> 1 12 1
> 2 1 0
> 2 2 0
> 2 3 0
> 2 4 1
> 2 5 1
> 2 6 1
> 2 7 0
> 2 8 0
> 2 9 0
> 2 10 2
> 2 11 2
> 2 12 0

*
*   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/