| Date: | Thu, 2 Jan 1997 17:39:22 -0700 |
|---|
| Reply-To: | "Raymond V. Liedka" <liedka@UNM.EDU> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU> |
|---|
| From: | "Raymond V. Liedka" <liedka@UNM.EDU> |
|---|
| Subject: | Re: COMPRESS Option (SAS 6.12) |
|---|
|
| In-Reply-To: | <m0vful6-0001NhC@pyxis.unm.edu> |
|---|
| Content-Type: | TEXT/PLAIN; charset=US-ASCII |
|---|
On Thu, 2 Jan 1997, Devi Katikineni wrote:
> Why would anyone NOT use the COMPRESS option in SAS?
>
If one needed to access the data by observation number. An example is if
you are using the point= syntax in a datastep. A compressed dataset
cannot be accessed except as a "whole"; that is you cannot jump from
recored to record (like using the point= sytax).
> He found that using OPTIONS COMPRESS significantly reduces both
> the process time and the dataset size. He expected that there
> would be some overhead with the COMPRESS option, and so process
> time would be longer, but found that it was shorter??
>
If the compression is pretty succesful and the original data is
voluminous, then you might save a lot of time reading a much smaller file
off the floppy drive. Hard drives are just plain speedier.
> Any comments would be welcome. TIA.
>
> Devi Katikineni
> Social & Scientific Systems, Bethesda, MD.
>
Raymond V. Liedka
Department of Sociology
University of New Mexico
| 
