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 (April 2004, week 5)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Thu, 29 Apr 2004 14:33:23 -0400
Reply-To:     Sigurd Hermansen <HERMANS1@WESTAT.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Sigurd Hermansen <HERMANS1@WESTAT.COM>
Subject:      Re: NO more than 32767 variables???
Comments: To: Aldi Kraja <aldi@WUBIOS.WUSTL.EDU>
Content-Type: text/plain

Aldi: Richard and Venky have already questioned why anyone would prefer to work with 60,000 variables. That's usually my line. I would go even farther and question why anyone needs 100 variables.

From your question I know you understand that transposing your data does not enhance the information the original dataset contains. A transpose merely replaces data in columns with variable labels. Have you considered using 'virtual arrays'?

Think of a row of data as ID, index value(s) and data element(s). Any operations that one can perform on a row of 60,000 variables, you can also perform on a virtual array. Basic tools include the SAS SQL GROUP BY and reflexive joins.

You may at some point need to transpose data from rows to columns that SAS statistical procedures specify. For repeated measures, maybe not .... I'll bet that you will not need 60,000 variables at that point. For SAS IML, perhaps, but then you could simply read into a data matrix and operate on a transpose.

I'll speculate that you can learn to handle virtual arrays in the time that it would take to transpose your data and print a data dictionary. Sig

-----Original Message----- From: Aldi Kraja [mailto:aldi@WUBIOS.WUSTL.EDU] Sent: Thursday, April 29, 2004 1:34 PM To: SAS-L@LISTSERV.UGA.EDU Subject: NO more than 32767 variables???

Hi,

I am transposing a dataset which has about 100 variables and about 60000 observations. By transpose I am changing the role of the data in the table: a. 100 variables will become observations and b. 60000 observations will become variables. All of a sudden I find that SAS does not permit more that 32767 variables in a table.

Any solution?

TIA, Aldi

ERROR: Too many variables defined for file IN1.SNPGANON2. This file may not have more than 32767 variables.^MERROR: Too many variables defined for file IN1.SNPGANON2. This file may not have more than 32767 variables.^MERROR: Too many variables defined for file IN1.SNPGANON2. This file may not have more than 32767 variables.


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