[SOLVED] Linux print column entries if entries for another column are equal

Issue

Using linux commands or bash, how can print the entries for column 6 as a single line when the corresponding entries for column 3 are the same value. e.g. in the sample file below, when column 3 has the value of 1 (rows 1 and 2) print '4 103', when column 3 takes the value of 2 (rows 3 and 4) print '113 4', and such on.

1  A  1  C  18   4
1  A  1  C  8    103
1  A  2  C  3    113
1  A  2  C  50   4
1  A  3  C  2    103
1  A  3  C  33   4
1  A  3  C  9    4
1  A  3  C  113  103
1  A  4  C  44   203

The expected output should be:

4 103 
113 4
103 4 4 103 
203

I have tried using awk for selecting the rows when column 3 takes a specific value (i.e. when column 3 is 1 only), but I cannot find a way to achieve this when column 3 can take any integer value. Thank you.

Solution

With any awk:

awk '
    BEGIN{last=1}
    $3 == last{printf "%d ", $6;last=$3;next}
    {last=$3; printf "\n%d ", $6}
    END{print ""}
' file

Output

4 103 
113 4 
103 4 4 103 
203 

Answered By - Gilles Quénot

Answer Checked By - David Goodson (WPSolving Volunteer)