Re: multiple samples w/ replacement from large datasets
- From: Michael.Lacy.junk@xxxxxxxxxxxxx
- Date: 17 Apr 2006 07:51:54 -0700
Data Matter <fungile@xxxxxxxxx> wrote:
The last step seems to require that the system can simultaneously build
all k samples. Lets say I'm creating 1000 samples. Can typical
programs like SAS and SPSS keep all 1000 samples in memory while I read
through the big data set?
I don't know about SAS. As for SPSS, I think a whole re-thinking
of the approach might be necessary. In SPSS, I think the best approach
would involve creating 1 variable with many values that could be
used with Split File. I suggest you repost your original question in
comp.soft-sys.stat.spss or the SAS group. As I indicated originally,
the solution to this problem will depend on what software you want to
use. The kind of solution I suggested might not be best
suited to SPSS.
--
=-=-=-=-=-=-=-=-=-==-=-=-=
Mike Lacy, Ft Collins CO 80523
Clean out the 'junk' to email me.
.
- References:
- multiple samples w/ replacement from large datasets
- From: Data Matter
- Re: multiple samples w/ replacement from large datasets
- From: Michael . Lacy . junk
- Re: multiple samples w/ replacement from large datasets
- From: Data Matter
- multiple samples w/ replacement from large datasets
- Prev by Date: Re: Covariance structure for random effect
- Next by Date: Re: interpreting significant interactions in Multiple Regression
- Previous by thread: Re: multiple samples w/ replacement from large datasets
- Next by thread: Hansen Jagannathan Distance
- Index(es):
Relevant Pages
|
