01-13-2017 09:45 AM
I currently have a dataset with 400 variables for 1,000,000 indivduals. I've been given a list of totals for certain variables however, the names are different to what's in my dataset.
Therefore, I would like to sum each of my variables in order to create a total for each variable, so going from 1,000,000 plus cases to just 1 for each varaible, with the 1 being the sum of all cases.
Each of the variables in my dataset have different names, but also there are character variables in it also.
Would anyone be able to suggest a work around for what I wish to do, or perhaps a simpler method to what I've proposed?
01-13-2017 10:00 AM - edited 01-13-2017 10:05 AM
Please post example test data - in the form of a datastep - so that we can see what we are working with. Only need a couple of lines.
It may be you can use the _numeric_ in a proc means or summary to get your required output, for instance:
proc summary data=sashelp.class;
output out=sumry sum=;
Edit, the above code should work.