Date:         Thu, 5 Jun 2008 22:57:53 +0530
Reply-To:     ohri2007@GMAIL.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         ohri2007@GMAIL.COM
Subject:      Re: Help
Comments: To: Raj Datta <dokiya777@gmail.com>
In-Reply-To:  <200806051712.m55AkqDv027229@malibu.cc.uga.edu>
Content-Type: text/plain; charset=ISO-8859-1

Use ;



Using

options compress =yes ;

will decrease size of dataset.

I would recommend indexing, but if it's not indexed already then it
will take a long time to create an index .
Simply use,

if cal <0 for creating the new dataset.

This may take some time.

Alternatively, some kind of hash method (I am not sure of ) will give
you a faster result.

Regards,

Ajay

On 6/5/08, Raj Datta <dokiya777@gmail.com> wrote:
> I am new to SAS/SQL and need some help. I have 10 million records dataset
> with multiple record id. Last column is the calcultation between row1-row2.
> Is it possible to extract rows in which there is a negative number in cal.
> for example in this example pick rows having -3,-3,-3.
>
> record id  amt Date Cal
> 1    10 Jan-08 1
> 1    9 Feb-08 1
> 1    8 Mar-08 -3
> 1    11 Apr-08 6
> 1    5 May-08 1
> 1    4 Jun-08 -3
> 2    7 Jan-08 1
> 2    6 Feb-08 1
> 2    5 Mar-08 -3
> 2    8 Apr-08 4
> 2   4 May-08 1
> 2    3 Jun-08 3
>
> Thanks in advance.
>
> RD
>