Date: Mon, 1 Sep 1997 08:52:44 -0400
Reply-To: Jules Bosch <0006974523@MCIMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: Jules Bosch <0006974523@MCIMAIL.COM>
Subject: Notes Re: Recent Posts
I finally have some time to add my input to some recent posts that have
caught my interest.
First, a SAS user wanted to know how to gather info about the variables in a
data set and the use of a macro was mentioned. A number of good solutions
followed. One I have used on occasion comes from the SAS Guide to Macro
Processing, Version 6, Second Ed., page 263. The macro provides the number
of observations in a data set. There are lots of explanatory notes with the
example so see the reference for more details.
It reads -
if 0 then set &dsn nobs=count;
Furthermore, the user was interested in discovering missing values in the
data set. The SAS product "SAS Today - A Year of SAS Tips" (a day-at-a-time
desk calendar full of interesting SAS info) lists a program that allows one
to get the count and % of missing values for each var in the given data set.
It reads -
. = 'missing'
._, .A-.Z = 'special missing';
' ' = 'missing'
other = 'not missing';
proc freq data=dsn;
tables _all_ / missing;
format _numeric_ NUMFMT.
Finally, there were comments re: the lenght of a macro. The 1Q97 issue of
"SAS Communications" has a technical tip that states the length of a macro
is 32K. An excellent example is included to demo the development of a macro
var which could easily have a length > 200 chars. The tip (p. 48) shows how
to create a list (a macro var) that can be used in a WHERE statement to
subset a master data set. The code from that tip follows.
input empl $;
input empl $ job $;
/* Create the macro var &LIST */
proc sql noprint;
select quote (trim(left(empl)))
into :list separated by " "
proc print data=master;
where empl in(&list);
After resolution of the macro var the WHERE statement looks like -
WHERE EMPL IN("12345" "67890" "45678"
"00001" "98765" ...);
and the values "12345" and "00001" are the only two strings found in the
master data set.
The technical tips in the "SAS Communications" are often invaluable.