As always, please remember for the edification
of others to quote the origin of Stata
commands that are user-written.
-makematrix- is available from SSC.
Without knowing more about your data, this
does appear strange. It may be that you
need to apply the -listwise- option.
With P-values as small as that, you don't
need these results anyway!
Nick
[email protected]
Herve STOLOWY
> I face the following strange phenomenon concerning the
> -ranksum- test and -makematrix-.
>
> If I use the following command:
>
> makematrix c1, from(r(z) 2*normprob(-abs(r(z)))) : ranksum
> exp_fi , by(interne)
>
> It works and I get the following matrix:
>
> c1[1,2]
> z 2*normprob(-abs(z))
> exp_fi 4.9154637 8.857e-07
>
> When I add two variables, my first variable is not disclosed anymore:
>
> makematrix c2, from(r(z) 2*normprob(-abs(r(z)))) : ranksum
> exp_fi fonction_ent fonction_exp, by(interne)
>
> c2[3,2]
> z 2*normprob(-abs(z))
> exp_fi . .
> fonction_ent -5.8967476 3.707e-09
> fonction_exp 5.8967476 3.707e-09
>
> I know that the last two variables are defined on the basis
> of the first one. But is this explanation sufficient to
> understand why the first variable "disappears"? Why couldn't
> I have the first variable also reported, as it is with the
> first command?
*
* 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/
