Shell Programming and Scripting

Hi,

I have a table like this:

The 1st column is just names of the rows. Columns A1, A2, A3 belongs to group A and columns B1, B2, B3 to B. I need to extract lines containing same element in the columns of each group or both groups.

So desired output is


Thanks in advance

You mean, rows where all the columns have something in common between each other? So kke would be related to ttk, but bvv wouldn't be related to cxx?

Try:

Code:

awk 'NR==1 || $2==$3 && $3==$4 && $5==$6 && $6==$7' infile

-or less robust-

Code:

sed -n '1p;/[^ ]*\( [^ ]*\)\1\1\( [^ ]*\)\2\2/p' infile

THanks for the reply. Sorry I was unable to explain properly in the first post.

The idea is that in each row, each group must contain three exactly same type of element. For example, one group should contain exactly same type of element in all three columns of a row (such as ttk ttk ttk in A ). It can not contain different elements (such as ttk ttk Ttk in A or cBx ccx ccx in B). In the same row, the other group must also contain same type of element in all the three columns (kke kke kke); Doesnt matter if its the same element as that of group A ( like mcr mcr mcr mcr mcr mcr).

Hope I havent confused you more

---------- Post updated at 05:10 PM ---------- Previous update was at 04:40 PM ----------

thanks a lot. you rock.