awk - Comparing files based on column -

i try write script comparing 2 large files based on column 2. each file contains 1 million records. output, need know records common on column 2 (exist on both files) have different value in column 1. files quoted comma separated value files

file1_pair  20151026,1111 20141113,2222 20130102,3333 77777777,9999  file2_pair 20151026,1111 20203344,2222 50506677,3333 77777777,8888  desired_output 20141113,2222,20203344 20130102,3333,50506677 

i tried modifying script below not able right.

awk 'fnr==nr { a[$0]; next } !($2) in { c++ } end { print c }' file1_pair file2_pair` 

you had right idea operating on wrong fields.

you need save $2 values first file in array , check $2 values second file against array. need compare value of $1 in corresponding rows.

this awk script that.

awk -f , -v ofs=, '     nr==fnr {         # store value of $1 under $2 key in         a[$2]=$1         next     }     # if $2 in (we've seen value before) ,     # if value in array (first file's $1 value) doesn't match files $1 value     ($2 in a) && (a[$2] != $1) {         # print original $1 value (from array),$2,$1         print a[$2],$2,$1     }' file1_pair file2_pair