04-18-2015 12:04 PM
I wonder how we add a vector column with different values in the same data-step?
One solution is to create a document with the vector then merge it with the original document.
But, I'm interested to do it in the same document. Any help is extremely appreciable.
04-18-2015 02:09 PM
Vectors and documents are not datastep concepts. As a rule, every variable created in a datastep becomes a variable (or column) in the dataset produced.
Please give an example (Input --> Output) of what you are trying to do.
04-19-2015 04:16 AM
Thanks for your prompt reply.
Here's the example: to create a document for a planned study that contains:
- two schools, 50 boys and 50 girls from each school - total 200 pupils.
- for each group (e.g. school 1, 50 boys) to randomly assign a number and sort. The first 25 pupils get treatment and the next 25 pupils get control.
What I've done is as follows:
do school=1,2; do gender='F','M'; do id=1 to 50;
proc sort; by school gender random;
data b; /*creating a vector of changing group values*/
do temp=1 to 4; do group='C','T'; do temp_1=1 to 25;
drop temp temp_1;
data all; merge a b;
This works fine, but I was asked to do it in one document without merging an external document (I was given a tip of using count+1).
Do you have an idea how to resolve this?