|
D.T.,
Just use the VLABEL function instead of VNAME. ;-)
Mike Rhoads
RhoadsM1@Westat.com
-----Original Message-----
From: D T [mailto:sasandstats@live.com]
Sent: Tuesday, May 12, 2009 2:36 PM
To: Mike Rhoads; sas-l@listserv.uga.edu
Subject: RE: [SAS-L] output question--one long row into many rows
Mike,
thanks for this suggestion! Yes, you were correct: since I have many variables, I would prefer a more automated approach, and yours worked perfectly. What occurred to me after I saw your code is this: all my variables starting with "num" actually have explanatory labels--is there a way you are aware of that would let me put the label text into the code variable?
I am thinking about this step:
code = substr(vname(_num[n]),4);
Thanks!
D.T.
> Date: Tue, 12 May 2009 13:38:06 -0400
> From: RHOADSM1@WESTAT.COM
> Subject: Re: [SAS-L] output question--one long row into many rows
> To: SAS-L@LISTSERV.UGA.EDU
>
> Since the original question indicates that there are many more variables than are listed in the example, you can streamline Mary's approach using arrays, the VNAME function (if a "code" variable is desired), and colon-style variable lists, assuming that the variables are systematically named and ordered as the question implies.
>
> data output_set;
> set set1;
> array _num {*} num: ;
> array _pct {*} pct: ;
> length code $ 8; /* or whatever length is needed */
> keep code num pct;
> format pct 5.1;
> if id=211 then do;
> do n = 1 to dim(_num);
> code = substr(vname(_num[n]),4);
> num = _num[n];
> pct = _pct[n];
> output;
> end;
> end;
> run;
>
> Mike Rhoads
> RhoadsM1@Westat.com
>
>
> -----Original Message-----
> From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Mary
> Sent: Tuesday, May 12, 2009 11:56 AM
> To: SAS-L@LISTSERV.UGA.EDU
> Subject: Re: output question--one long row into many rows
>
>
> Try something like this (NOT tested)
>
> data output_set;
> set set1;
> if id=211 then
> do;
> code='a';
> num=numa;
> pct=pcta
> output;
> code='lt';
> num=numlt;
> pct=pctlt;
> output;
> num=numo;
> pct=pcto;
> output;
> code='f';
> num=numf;
> pct=pctf;
> output;
> end;
> keep id code num pct;
> run;
>
>
> ----- Original Message -----
> From: "D T" <sasandstats@LIVE.COM>
> To: <SAS-L@LISTSERV.UGA.EDU>
> Sent: Tuesday, May 12, 2009 9:51 AM
> Subject: output question--one long row into many rows
>
>
> I have a data set like this (I am just showing some
> variables, there are more):
>
> ID numa pcta numlt pctlt numo pcto numf
> pctf
>
> 1 52 2.6 13 0.6 6 0.3 220
> 10.9
>
> 2 13 0.6 101 5.0 11 2.1 13
> 0.7
>
> [...]
>
> 211 905 4.0 732 3.6 87 3.3 1176
> 9.8
>
>
>
> I need to output data for ID 211 only-this is a summary row.
> I can use a proc print and get the information I need printed (one long row of output), but am wondering if there isn't way I can make the output more user-friendly.
>
> Here is what I would like to have in a table, if possible:
>
> numa pcta
>
> numlt pctlt
>
> numo pcto
>
> numf pctf
>
> .... etc.
>
>
>
> In terms of output, it would look like this for my example:
>
> 905 4.0
>
> 732 3.6
>
> 87 3.3
>
> 1176 9.8
>
>
>
> Can anyone think of an efficient way to code this output?
>
>
>
> Thanks!
>
> D.T.
>
>
> _________________________________________________________________
> Hotmail(r) has ever-growing storage! Don't worry about storage limits. http://windowslive.com/Tutorial/Hotmail/Storage?ocid=TXT_TAGLM_WL_HM_Tutorial_Storage1_052009=
________________________________
Windows Live(tm): Keep your life in sync. Check it out.<http://windowslive.com/explore?ocid=TXT_TAGLM_BR_life_in_synch_052009>
|
