Date:         Thu, 3 Jan 2008 17:08:24 -0500
Reply-To:     Mike Rhoads <RHOADSM1@WESTAT.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mike Rhoads <RHOADSM1@WESTAT.COM>
Subject:      Re: creating 1 hour averages
Comments: To: firepit <FirePitMarta@gmail.com>
In-Reply-To:  <3b8837c0-12db-4191-af74-d7cb22665143@l1g2000hsa.googlegroups.com>
Content-Type: text/plain; charset="us-ascii"

There have been several good responses already, including Dan's
cautions.

It may be that your data are already in a SAS data set, with TIME as a
"time" variable.  If so, you can format that variable within your PROC
SUMMARY and avoid an extra step as follows (adapted from Nat's example):

/* Assume this step was run previously */
data avg;
informat time time.;
format   time time.;
input Year    Month   Day     Time    Speed;
* hour = hour(time);
cards;
2007    10      30      14:13   3.7
2007    10      30      14:14   5.7
2007    10      30      14:15   2.2
2007    10      30      14:16   3.7
2007    10      31      02:27   2.2
2007    10      31      03:15   3.3
run;

proc summary data = avg nway;
output out = avg mean = avg_speed;
var speed;
by year month day time;
format time hour2.;
run;

proc print;


Mike Rhoads
Westat
RhoadsM1@Westat.com

-----Original Message-----
From: owner-sas-l@listserv.uga.edu [mailto:owner-sas-l@listserv.uga.edu]
On Behalf Of firepit
Sent: Thursday, January 03, 2008 2:42 PM
To: sas-l@uga.edu
Subject: creating 1 hour averages


I am trying to create a 1 hour average for speed  from data that looks
like:

Year    Month   Day     Time    Speed
2007    10      30      14:13   3.7
2007    10      30      14:14   5.7
2007    10      30      14:15   2.2
2007    10      30      14:16   3.7

Google hasn't been very helpful...