Statalist


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

st: RE: Re: re: counter variable


From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   st: RE: Re: re: counter variable
Date   Sat, 2 May 2009 17:18:31 +0100

If the data have been -tsset id schlyr- and -tsspell- has been
downloaded from SSC then 

tsspell, fcond(newprinc == 1) 

is enough. This creates three new variables, of which _seq is the
counter desired. 

Nick 
[email protected] 

Martin Weiss

You have to hand it to all contributors here, though, that this is not
your 
everyday, apply-bysort-straightforwardly-and-look-at-Nick`s-column- 
in -the-SJ-carefully problem (see 
http://www.stata-journal.com/sjpdf.html?articlenum=pr0004). The fact
that 
you want this thing sorted, but not counted, by schoolyear and counted,
but 
not sorted, by the new dummy, makes this quite involved, and despite my
best 
effort, I cannot find a one liner for it that does not make extensive
use 
of -cond-. So kudos to Eva for her two liner :-)


"Kit Baum" <[email protected]>

><> T suggested
>
>  g newvar = 1 if new_prc==1
>  bys Id (schlyr): replace newvar = newvar[_n-1]+1 if newvar !=1
>
> This sounds to me like a generally bad idea. The first statement just 
> changes the new_prc zero values into missing values. The second
statement 
> relies upon the fact that missing is a value not equal to 1.  Although
the 
> logic works, I think Eva's suggested approach is a better  way to
think 
> about a problem like this (although I deplore her  reliance on -set 
> varabbrev on-, which is IMHO a Very Bad Idea).

*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index