No one has yet replied to this, perhaps for
the same reason that I nearly didn't.
I see here
(a) some data
(b) some code
(c) a question: "Is this the best approach?"
I can't answer (c) on the basis of (a) and
(b) without knowing what you are trying
to do. Perhaps that should be clear to me,
but it isn't.
Nick
[email protected]
Rajesh Tharyan
>
> Is there a way to create groups (say 25 groups) by two
> variables? Here X and
> Y,
>
> Year id X Y
> 986 6215 1 .645
> 1986 1781 1 .778
> 1986 2774 1 .819
> 1986 6354 1 .921
< snip>
> 1986 433 3271 .9935799
> 1986 753 4617 .7927225
> 1986 2675 4930 .7089249
> 1986 817 10212 .962593
>
> I can create, quintiles of X and then for each of those
> create quintiles of
> Y giving me 25 groups.
>
> Using the following code
>
> sort X
> xtile mc=X,nq(5)
>
> gen bm=.
> levels mc, local(levels)
> foreach i of local levels {
> xtile bm_`i'=Y if mc==`i',nq(5)
> replace bm=bm_`i' if mc==`i'
> drop bm_`i'
> }
> tostring bm mc,replace
> gen a=mc+bm
>
> I want to see how this differs from a grouping based on a
> double sort of X
> and Y.
>
> So I need to do
>
> sort x y
> gen b=_n
> gen group=floor(_n/10) * as in this case I have 264 observations)
>
> is this the best approach?
*
* 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/
