How to remove lines according a column?

fantasy1215 · 06-12-2014, 03:29 AM

I have a file like this:
time case_num other_columns

Code:

20140612162021    3000               others
20140612162022    3001               others
20140612162023    3001               others
20140612162024    3002               others
20140612162025    3002               others
20140612162026    3002               others

According case_num, when the same case_num has more than one line, I'd like to leave such line which has the most recently time.

after remove lines the result file is like

Code:

20140612162021    3000               others
20140612162023    3001               others
20140612162026    3002               others

I'd appreciate your help, many thanks.

grail · 06-12-2014, 03:34 AM

And what have you done to try and solve this problem? If you are not sure what tool to look at, I would suggest awk as it is designed with columned data in mind

danielbmartin · 06-12-2014, 08:04 AM

This is clumsy but it works.

With this InFile ...

Code:

20140612162021    3000               others
20140612162022    3001               others
20140612162023    3001               others
20140612162024    3002               others
20140612162025    3002               others
20140612162026    3002               others

... this code ...

Code:

tac $InFile |awk '!a[$2]++' |tac >$OutFile

... produced this OutFile ...

Code:

20140612162021    3000               others
20140612162023    3001               others
20140612162026    3002               others

Daniel B. Martin