Date:         Wed, 21 May 2003 11:19:52 -0400
Reply-To:     "Cacialli, Doug" <Doug_Cacialli@URMC.ROCHESTER.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Cacialli, Doug" <Doug_Cacialli@URMC.ROCHESTER.EDU>
Subject:      Any suggestions?
Content-Type: text/plain

Morning everyone!

I'm hoping that some of you can provide me with some criticisms and/or
suggestions about some code I've written.

The following code is part of a macro.  One of the things the macro is
designed to do is subset a data set on first or last observation for every
value of a given variable.  Of course to do this, I need to specify how the
data set is sorted.  I could specify a parameter to input the sort
variable(s), but I'd like to automate this as much as possible.  So what
I've done is:

PROC CONTENTS DATA = &INPUTDS._SORTED OUT = &INPUTDS._CONTENTS NOPRINT;
RUN;
PROC TRANSPOSE DATA = &INPUTDS._CONTENTS OUT = TRANSPOSED PREFIX = SORT;
WHERE SORTEDBY ^= .;
BY MEMNAME;
ID SORTEDBY;
VAR NAME;
RUN;
DATA SORTVARS;
        SET TRANSPOSED;
ARRAY NSORTVARS (*) SORT: ;
NVARS = DIM(NSORTVARS);
        IF              (NVARS = 1)
        THEN    CONCATENATED = TRIM(LEFT(SORT1));
        ELSE IF (NVARS = 2)
        THEN    CONCATENATED = TRIM(LEFT(SORT1)) || ' ' ||
TRIM(LEFT(SORT2));
        ELSE IF (NVARS = 3)
        THEN    CONCATENATED = TRIM(LEFT(SORT1)) || ' ' || TRIM(LEFT(SORT2))
||      ' ' || TRIM(LEFT(SORT3));
        ELSE IF (NVARS = 4)
        THEN    CONCATENATED = TRIM(LEFT(SORT1)) || ' ' || TRIM(LEFT(SORT2))
||      ' ' || TRIM(LEFT(SORT3)) || ' ' || TRIM(LEFT(SORT4));
CALL SYMPUT ('SRTVARS',CONCATENATED);
RUN;

There will never be more than four sort variables. The input data set will
always be sorted, and of the form '[DATA SET NAME]_SORTED'.  &INPUTDS is a
parameter of the macro.  After &SRTVARS is created by CALL SYMPUT, it is
used later in the BY statement of the data set where the subsetting takes
place.

Does anyone see any obvious ways that this can be made more efficient or
less cumbersome?  Thanks for any thoughts.

Doug out.

---------------------------------------------------
Douglas Cacialli - Data Manager / Data Analyst
Sleep and Depression Research Laboratory
University of Rochester Medical Center
300 Crittenden Blvd. - Box PSYCH
Rochester, New York 14642
Phone: (585)273-3309          Fax: (585)506-0287
---------------------------------------------------