07-12-2017 10:46 AM
Dear SAS communities
I have a excel dataset that contains a 50 binary variables (e.g. history of diabetes: 1/0 etc), a variable "center" which contains the name of one of 10 hospitals
I just want to create a summary stats excel table where each of the 50 binary variables is listed as a line like this
variable N. Nmiss. frequent "0" percentage "0" frequent "0" percentage "0"
hx_diabetes. xx. xx%. yy. yy%.
hx_stroke xx. xx%. yy. yy%.
then sort by "center" to output the same table restricting to each of the "center"
if you have a code that I can borrow, I would be grateful. Thanks
07-12-2017 10:59 AM
Your last variables are duplicated so not sure what you're looking for there. Otherwise this can be standard output from PROC MEANS.
PROC MEANS can do the N, NMISS. Frequency of a 1/0 variable is the SUM and Percentage is the Mean.
Note that you'll get a value for the Overall as well as each facility with this code.
proc means data=have out=want stackods n nmiss sum mean; class hospital; var hx_: ; *use all hx variables; ods output summary=MySummary; run;