|
PROC MIANALYZE on PROC MEANS result
Thanks Rob. [SAS 7610176377] proc mianalyze
data missing;
do rep=1 to 100;
if ranuni(12143)>.4 then x=rannor(2111)*100+44; if ranuni(13330)>.9
then y=rannor(2211)*10; if ranuni(12213)>.5 then z=rannor(2114)+70; output; end;
ods trace on;
proc mi data=missing out=impute nimpute=3 seed=1000; ods select parameterestimates; var
x y z; run;
proc means data=impute mean stderr noprint; var x y z; by _imputation_; output out=mean_ds mean=mean_x mean_y
mean_z stderr=stderr_x stderr_y stderr_z; run;
proc mianalyze data=mean_ds;
ods select parameterestimates;
modeleffects mean_x mean_y mean_z;
stderr
stderr_x stderr_y stderr_z;
run;
Length
Thanks, Jan. Based on [SAS 7610138678] question about the length of a numeric variable
data test;
input numval :;
format numval best20.;
datalines;
2.2
1
5.43243
10003.300204
;
data result;
set test;
digit_count = length(strip(put(numval,best20.)));
run;
proc print;
run;
Adding leading zeros (e.g., 123 --> 0123)
(Thanks, Kim. Based on [SAS 7610121586] A question about a character function)
In order to create variable NEW_ID, you can use the PUT function and the format Zw. which adds leading zeros to the specified
length. Here's an example:
new_id=put(id,z4.);
Sample 26140: Creating a new data set for each BY-Group in a data set
Thanks, Russ. [SAS 7610122275] break a data into separate pieces by a group ID
|
