SAS Two-way count table: How to produce?

Question

I have been using the proc freq command to create a two-way table of counts. I have two columns of variables, and want to make a matrix, such that the variables in Column A are grouped in rows, and the variables in Column B become columns. The names of the columns are the variables from Column B. How does one produce a SAS dataset like this, where the entries in the table are the counts of the Column A/Column B pairings?

Answer 1

Don,

If I understand your question correctly then the example below might work. There will be an issue in that combinations of column_a and column_b that do not exist will show up as missing values, but you could redefine them to zero in a data set afterwards if that matters.

proc freq data=dsn_in;
   tables column_a * column_b /
      out=dsn_out;

proc sort data=dsn_out;
   by column_a column_b;
proc transpose data=dsn_out out=dsn_transpose(drop=_label_ _name_);
   by column_a;
   id column_b;
   var count;
run;