That's droll, but I imagine that people interested in comparing three or
more files really would want to see some kind of Venn diagram, or
equivalent. That's not the same as all the pairwise comparisons. I am
not sure that the effort of programming that really nicely corresponds
with the amount of use it would get, however.
Nick
[email protected]
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Martin Weiss
We can change that with your very own -ssc d tuples-, though....
*************
sysuse auto, clear
drop mpg
save auto100, replace
drop gear
save auto101, replace
drop trunk
save auto102, replace
tuples auto auto100 auto101 auto102, max(2)
loc omitted 0
forval i = 1/`ntuples' {
loc mycount: word count `tuple`i''
if `mycount' ==2{
di _n in red "Now we are comparing `tuple`i''"
cfvars `tuple`i''
}
else{
loc ++omitted
}
}
di in red "Omitted `omitted' tuples"
Nick Cox
No; not all, as (e.g.) Austin touched on the question of comparing two
_or more_ datasets, not something I wanted to do.
Martin Weiss
Sorry about that! Let me check whether Nick has superseded all our
discussions yesterday with his new package...
*
* 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/
