Re: st: Re: new sampling problem

["Michael Blasnik" <michael.blasnik@verizon.net>]

...
I don't really understand what you are doing here, but the most obvious error 
is -local news=newvar1-.  Where is newvar1 defined?  If newvar1 is a variable, 
then the local macro will only contain the first observation of that variable 
because you can't assign a variable to a macro like that.  Your code is too 
convoluted for me to figure out how to fix it.  I'm also fairly certain that you 
don't want to use mkmat, restore, svmat in sequence.

Michael Blasnik

----- Original Message ----- 
From: "John allfish" <j_w11111@hotmail.co.uk>
To: <statalist@hsphsun2.harvard.edu>
Sent: Tuesday, June 19, 2007 12:28 PM
Subject: Re: st: Re: new sampling problem



> Thanks for this, and for the setting the seed advice.
> I have tried to sort this out with this very crude program but it only seems 
> to do the sampling for one value of newvar1, can anyone help me with this? the 
> first part of the do file defines the data. The last part does the sampling I 
> am interested in. It is very, very crude - and it may seem a little odd, 
> although I know what I want to do.
>
> clear
> range y 1 10 10
> sum y
> gen income=.
> replace income=1 in 1/4
> replace income=2 in 5/8
> replace income=3 in 9/10
> gen sex1=.
> replace sex1=2 in 1/5
> replace sex1=1 in 6/10
> gen sex2=.
> replace sex2=1 in 1/3
> replace sex2=2 in 4/5
> replace sex2=3 in 6/10
> tab income
> tab sex1
> tab sex2
>
> preserve
> sort sex2
> contract sex2
> tempvar newvar
> samplepps newvar, size(_freq) withrepl ncases(100)
> mkmat newvar
> restore
> svmat newvar
> gen const=1
>
> local news=newvar1
> foreach val of local news{
> samplepps sample`val', ncases(`val') withrepl size(const)
> }
<snip> 
*
*   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/