grep -v ",0$" dataset.dat | awk -F',' '{print $2}' | sort | uniq foo2 foo3 foo4