If I have a character variable with the value "45 000" and want to turn it into a numeric variable, the space is obviously going to cause a problem. Can you please tell me how I can edit my expression to resolve this?
input(varname, 8.0)
Incidentally, can you also please advise on how I can make my data step exclude the last row from the data set? The total number of rows will change from month to month.
Thanks.
For the first part of your question...I'm not familiar with DI Studio but this works in EG 4.3...maybe it'll help:
DATA grades;
input idno 1-2 variable $ 5-11;
cards;
01 45 000
02 32 000
03 100 000
;
RUN;
data grades_1(keep=idno variable2);
set grades;
variable1=CATS(SCAN(variable,1), SCAN(variable,2));
variable2=INPUT(variable1, best.);
run;
PROC PRINT data = grades_1 NOOBS;
title 'The grades data set';
RUN;
As for the last row, I think the simplest is to have a data step in a user written transform, in which you specify if end=last then delete; or something similar.
The comma informat should do the trick.
try:
data want;
set yourdataset end=last;
newvariable=input(compress(varname),8.);
if not last;
run;
hi ... the COMMA informat removes internal blanks ...
data grades;
input x $20.;
datalines;
45 000
32 000
1 2 3 4 5
1 0
100 000
;
data grades;
set grades end=last;
y = input(x,comma20.);
if ^last;
run;
x y
45 000 45000
32 000 32000
1 2 3 4 5 12345
1 0 10
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
Need to connect to databases in SAS Viya? SAS’ David Ghan shows you two methods – via SAS/ACCESS LIBNAME and SAS Data Connector SASLIBS – in this video.
Find more tutorials on the SAS Users YouTube channel.