I did read your entire post, twice. In
your original mail you said nothing about your data.
Giving output is one thing, but the data are another.
There is a long history of people claiming bugs
on the basis of little or no evidence that can
be discussed. It is good advice not to claim
a bug unless you can identify the code segment
that is problematic. This does not require
a reporter to be a Stata programmer. Sometimes
users can -set trace on- and see where code
stops with an error.
This reply mail is concrete and lets someone say something
more helpful.
At this moment I am on a computer away from base which
has just Stata 8.2. I get different output.
. tabodds foreign rep78, or cornfield
--------------------------------------------------------------------------
| odds Cornfield
rep78 | cases controls ratio [95% Conf. Interval]
------------+-------------------------------------------------------------
1 | 0 2 . . .
2 | 0 8 . . .
3 | 3 27 . 0.04267 .
4 | 9 9 . 0.43332 .
5 | 9 2 . 1.51393 .
--------------------------------------------------------------------------
Test of homogeneity (equal odds): chi2(4) = 26.87
Pr>chi2 = 0.0000
Score test for trend of odds: chi2(1) = 23.85
Pr>chi2 = 0.0000
. which tabodds
*! version 6.2.6 22apr2003
Does your problem arise with a later or earlier version of
-tabodds-?
Nick
[email protected]
Ricardo Ovaldia
--- Nick Cox <[email protected]> wrote:
> You need to say more about the data. Otherwise
> "This seems to be a bug" means nothing except
> "These results surprise me".
Please read the entire post before replying. I pasted
the output, both with and without the option
specified, to my original email. I Though that was
clear. Here is another example using the auto data and
commands:
. sysuse auto, clear
. tabodds foreign rep78
. tabodds foreign rep78, or cornfield
. sysuse auto, clear
(1978 Automobile Data)
. tabodds foreign rep78
--------------------------------------------------------------------------
rep78 | cases controls odds
[95% Conf. Interval]
------------+-------------------------------------------------------------
1 | 0 2 0.00000
. .
2 | 0 8 0.00000
. .
3 | 3 27 0.11111
0.03371 0.36625
4 | 9 9 1.00000
0.39695 2.51919
5 | 9 2 4.50000
0.97229 20.82703
--------------------------------------------------------------------------
Test of homogeneity (equal odds): chi2(4) = 26.87
Pr>chi2 = 0.0000
Score test for trend of odds: chi2(1) = 23.85
Pr>chi2 = 0.0000
. tabodds foreign rep78, or cornfield
--------------------------------------------------------------------------
| odds
Cornfield
rep78 | cases controls ratio
[95% Conf. Interval]
------------+-------------------------------------------------------------
'.' found where integer expected
r(7);
Ricardo Ovaldia
> > I get an error when I request cornfield CI in
> > -tabodds-. This seems to be a bug. Can anyone
> please
> > verify. Thank you.
> >
> > Ricardo.
> >
> > . tabodds vitb1p1 snp if case==1, or
> >
> >
>
--------------------------------------------------------------
> > -------------
> > snp | Odds Ratio chi2 P>chi2
>
> > [95% Conf. Interval]
> >
>
-------------+------------------------------------------------
> > -------------
> > CC | 1.000000 . .
>
> > . .
> > CT | 0.901799 0.26 0.6086
>
> > 0.607030 1.339707
> > TT | 0.663223 2.01 0.1564
>
> > 0.374388 1.174892
> >
>
--------------------------------------------------------------
> > -------------
> > Test of homogeneity (equal odds): chi2(2) =
> 2.02
> > Pr>chi2 =
> 0.3638
> >
> > Score test for trend of odds: chi2(1) =
> 1.79
> > Pr>chi2 =
> 0.1811
> >
> > . tabodds vitb1p1 snp if case==1, or cornf
> >
> >
>
--------------------------------------------------------------
> > ------------
> > | odds
>
> > Cornfield
> > snp | cases controls ratio
>
> > [95% Conf. Interval]
> >
>
------------+-------------------------------------------------
> > ------------
> > '.' found where integer expected
> > r(7);
>
*
* 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/
