Re: st: Re: How to merge individual records to groups in a largedataset, w/o using collapse

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

By levels, I mean distinct values, as in what the -levelsof- command 
provides in Stata.  I can't imagine what you think I meant -- if it had just 
one level then it would be a constant?

Anyway, a key issue for designing the best approach in terms of memory and 
speed requires knowing what a tabulation of region6 would look like..how 
many distinct groups does it define?  How big is the largest group? etc..


----- Original Message ----- 
From: "Allon Crazy" <allon_crazy@yahoo.com>
To: <statalist@hsphsun2.harvard.edu>
Sent: Tuesday, February 27, 2007 7:23 PM
Subject: Re: st: Re: How to merge individual records to groups in a large 
dataset, w/o using collapse



> Hi Michael,
>
> Thanks for your patience.
>
> My Region6 has only one level, which is why I did not
> answer directly. I was not sure what do you mean by
> the levels of Region6. Region6 has only one level, but
> many values (such as all the fips codes of counties).
>
> --- Michael Blasnik <michael.blasnik@verizon.net>
> wrote:
>
> > I think you misunderstood me.  I didn't suggest
> > doing one variable at a
> > time, although that could be built into the loop too
> > and may be a viable
> > alternative overall.  What I suggested was to
> > analyze  one category  of
> > group6 at a time within a loop and only using only
> > the variables in the
> > dataset that you need summary stats for (assuming
> > rhere may be other
> > variables in that large dataset that aren't needed
> > for this task) .  The
> > entire process can be written into a do or ado file
> > to automatically create
> > the results dataset you want.  You didn't bother
> > answering my questions
> > about how many levels there are in region6 and
> > whether the variables you
> > need could fit in memory using the largest region6
> > category, so I won't
> > bother showing you the code to automate it :)
> >
> > Michael
*
*   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/