LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (October 2004, week 2)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:   Tue, 12 Oct 2004 12:13:50 -0700
Reply-To:   "Schwarz, Barry A" <barry.a.schwarz@BOEING.COM>
Sender:   "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:   "Schwarz, Barry A" <barry.a.schwarz@BOEING.COM>
Subject:   Re: data mining
Content-Type:   text/plain; charset="us-ascii"

Try something like the following

DATA new (KEEP= location within incensus numfam);

/* read one location and determine if numfam is equal to count of incensus families */

DO UNTIL(LAST.address); SET old (RENAME location=address); BY address; IF FIRST.address THEN count = numfam; ELSE DO; IF incensus = 'Y' THEN count + -1; END; END;

/* re-read that location and generate output if not equal */

DO UNTIL(LAST.location); SET old; BY location; IF count ^= 0 THEN OUTPUT; END; RUN;

-----Original Message----- From: Daniel Reyes [mailto:junkdalejreyes@yahoo.com] Sent: Tuesday, October 12, 2004 5:14 AM To: Schwarz, Barry A Subject: RE: data mining

Oop!...darn....101 should have had a 3 NUMFAM value to highlight an example of what I don't want. Since NUMFAM is three and there are only two records with "Y" flags, then this would not be included in the output.

And yes the files are SAS datasets.

dan

--- "Schwarz, Barry A" <barry.a.schwarz@boeing.com> wrote:

> Is your data in a SAS data set or in a sequential > file. If the latter, > is it sorted as implied by your sample? > > Should not 101 Main St be included in your sample > output since NUMFAM is > 2 but only 1 has Y? > > Barry Schwarz > OS/390 System Programmer > M/S 80-JE > Phone: 253-773-4221 > Fax: 253-773-1257 > > > -----Original Message----- > From: Daniel Reyes [mailto:junkdalejreyes@YAHOO.COM] > > Sent: Thursday, October 07, 2004 10:40 AM > To: SAS-L@LISTSERV.UGA.EDU > Subject: data mining > > > I have a dataset of what we call basic street addresses--locations in > which more than one address may be found, like > apartment buildings. > Examples > include: > > 101 Main St Y 2 > 101 Main St Apt1 Y 0 > 101 Main St 2FL N 0 > > 102 Main St Y 3 > 102 Main St 2 N 0 > 102 Main St 3rd FL N 0 > 102 Main St BSMT N 0 > > 103 Main St Y 4 > 103 Main St #4 N 0 > 103 Main St STE6 Y 0 > 103 Main St ATTC Y 0 > > Each record of the BSA will have a > "within-structure" identifier (eg, 1 > or BSMT, etc), EXCEPT for the first instance when > sorted by the address. > > Each record has a flag set called INCENSUS (Y or N). > > Each record has a flag set called NUMFAM (numeric, > 0-999). > > I want to extract all BSAs (and their corresponding > records) in which the number of the NUMFAM flag DOES > NOT equal the total of good census records for the > BSA > (Ys). > > Therefore, for the example above, I ideally would > like > to see this in the output: > > 102 Main St Y 3 > 102 Main St 2 N 0 > 102 Main St 3rd FL N 0 > 102 Main St BSMT N 0 > > 103 Main St Y 4 > 103 Main St #4 N 0 > 103 Main St STE6 Y 0 > 103 Main St ATTC Y 0 > > Hope someone can help...please email me directly > > Thanks. > Dan > > > > __________________________________ > Do you Yahoo!? > Y! Messenger - Communicate in real time. Download > now. > http://messenger.yahoo.com >

__________________________________ Do you Yahoo!? Yahoo! Mail Address AutoComplete - You start. We finish. http://promotions.yahoo.com/new_mail


Back to: Top of message | Previous page | Main SAS-L page