Re: st: Collecting Statistics of Averages of Variables

[robert hartman <rohartman@gmail.com>]

But that is a good point, Clyde. I would probably look to cap the
number, so the best 2, 3, or 10-variable average out of the full set.
I'd have been better off leaving my actual number of variables out of
the original note. Apologies, and good point.

On Sat, Sep 29, 2012 at 5:35 PM, robert hartman <rohartman@gmail.com> wrote:
> By my lights, the number of unique combinations of 2 variables from a
> pool of 70 variables is 2415. That's a lot but not intractable
>
> On Sat, Sep 29, 2012 at 12:39 PM, Clyde B Schechter
> <clyde.schechter@einstein.yu.edu> wrote:
>> Robert Hartman asked about calculating summary statistics from all possible combinations of some variables.  The example in his initial post involved 4 variables, which should be no problem at all.  But at the end of his post he says:
>>
>> "I have about a 70 variable space from which to create all these subset combinations."
>>
>> If that means that he wants to do this for all possible combinations of 70 variables, that is 2^70 (= 1.181e+21) combinations to work out.  That's beyond merely time consuming: if each combination can be completely processed in 1 microsecond, we are looking at over 37,000,000 years of processing.  It's examples like this that evoke the phrase "combinatorial explosion."
>>
>> Perhaps the whole approach needs to be reconsidered.
>>
>> Clyde Schechter
>> Dept. of Family & Social Medicine
>> Albert Einstein College of Medicine
>> Bronx, NY, USA
>>
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/