Thank you, Martin and Howard. I guess the main thing I learned is how
to define and invoke local variables. With no background in
programming, I'm still struggling with how to loop a procedure. Good
thing copy and paste exist.
On Mon, Jul 6, 2009 at 6:14 PM, Martin Weiss<[email protected]> wrote:
>
> <>
>
> So here is the code w/o line delimiters as I suspect they do cause problems
> for some list users. Beginners might also be disappointed when pasting these
> line delimiters interactively and getting error messages...
>
>
> *************
> clear*
> set obs 10000
>
> gen x=rnormal()
> centile x, ce(10)
> local p10x=`r(c_1)'
>
> gen z=rnormal()
> centile z, ce(10)
> local p10z=`r(c_1)'
>
> * get indicator y
> gen byte y=!(x<=`p10x' & z <= `p10z')
> replace y=2 if x>`p10x' & z < `p10z'
> replace y=3 if x>`p10x' & z > `p10z'
>
> * check
> su x z, det
> table y, c(min x max x min z max z freq) row
> *************
>
>
>
> HTH
> Martin
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Pancho Villa
> Gesendet: Montag, 6. Juli 2009 22:32
> An: [email protected]
> Betreff: Re: st: RE: Converting a continuous var into a binary var
>
> On Mon, Jul 6, 2009 at 4:15 PM, Martin Weiss<[email protected]> wrote:
>>
>> <>
>>
>>
>> Am I right that for y=2 the z value does not matter at all? That is what I
>> take away from your description of the problem...
>
> Oops, the last line was supposed to be:
>
> y = 3 if x > 10th percentile of x & z > 10th percentile of z
> *
> * 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/
>
>
> *
> * 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/
>
*
* 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/
