Re: st: error message r(2000)

[Roger Harbord <roger.harbord@bristol.ac.uk>]

Are you running the -count- command inside the do-file which defines 
`varlist'? If not (e.g. you're running it from the command-line) then 
`varlist' will not be defined as it's local to the do-file, and missing() 
with no args is always true. If `varlist' contains >1 variable then 
missing(`varlist') shouldn't in fact work as missing() takes a 
comma-separated list not a varlist:

. local varlist  price mpg rep78 headroom

. di "`varlist'"
price mpg rep78 headroom

. count if ~missing(`varlist')
pricempgrep78headroom not found
r(111);

. count if ~missing()
  74


Roger.

--
Roger Harbord
Department of Social Medicine, University of Bristol
http://www.epi.bris.ac.uk


--On 31 March 2005 13:14 +0100 Ron�n Conroy <rconroy@rcsi.ie> wrote:


> louis boakye-yiadom wrote:
>
> > Dear All,
> > I get a "no observations" error message -r(2000)- when I execute a do
> > file,
> > and I'm having difficulty determining what's going wrong. According to
> > the
> > error message, I "have requested some statistical calculation and
> > there are
> > no observations on which to perform it. Perhaps you specified if exp
> > or in
> > range and inadvertently filtered all the data." I don't understand
> > what is
> > meant by "filtered all the data". The line that generates the error
> > message
> > is:
> > regress lnW `varlist' if migst1==0
> > To check whether there are actually no observations, I gave the following
> > command (and had the shown output):
> > count if migst1==0 & !mi(`varlist')
> > 2720
> Is there a non-numeric variable in there perhaps?
>
> --

*
*   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/