Date:         Fri, 11 Sep 1998 01:34:51 GMT
Reply-To:     shiling@MATH.WAYNE.EDU
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Shiling Zhang <shiling@MATH.WAYNE.EDU>
Organization: Deja News - The Leader in Internet Discussion
Subject:      Re: Output based on Input

James,
 
transpose procedure seems very simple. After you have the output data, a
looping will generate a set of datasets you want.
 
hope this helps.
 
proc transpose data=t1 out=t2(drop=_name_) prefix=L;
  by field1 field2;
  id field4;
  var field5;
run;
proc print; run;
 
output file;
****************;
                     OBS    FIELD1    FIELD2    L10001    L10002    L10003
 
                      1      88888     67789     11111     22222     33333
                      2      99999     66999     44444     55555         .
 
 
 
 
In article <Ez1pKB.EMy@rc.gc.ca>,
  "James Bennett" <bennettj@rc.gc.ca> wrote:
> Hello. Does anybody have an elegant solution to the following problem:
> I have data to be read in from a flat file and stored in different datasets
> based on a few of the fields read in.
> For example:
> Sample data to be read in:
> Field1    Field2    Field3    Field4    Field5
> 88888    67789    0100    10001    11111
> 88888    67789    0100    10002    22222
> 88888    67789    0100    10003    33333
> 99999    66999    0101    10001    44444
> 99999    66999    0101    10002    55555
>
> I need to read in each row one at a time and based on the value in Field3
> store the remaining fields in already existing datasets named "T"&Field3
> (ie:T0100 - rows 1 to 3 OR T0101 - rows 4 to 5). In addition Field4
> determines the field name in the dataset (ie: 10001 is field L10001).
> The result looks like the following:
>
> Dataset - T0100
> Field1    Field2    L10001    L10002    L10003
> 88888    67789    11111     22222      33333
>
> Dataset - T0101
> Field1    Field2    L10001    L10002
> 88888    67789    44444     55555
>
> I thought about constructing a dataset with a series of SQL Insert
> statements but I am not sure this is the way to go.
>
> *** Any help is very much appreciated. ***
>
>
 
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp   Create Your Own Free Member Forum