Shell Programming and Scripting

Dear Rudic

I did not bring the whole data set of file1 and file2 here. That's what you think these two file do not match together. The question is file2 have unique number of id in the second column but in file1 there could be repeated time of ids in column first. I want to know why after merging these two file, I do not get the same line in file1 which it had before.

Applying my proposal in post#6 to the above files yields

Code:

awk 'FNR==NR {T[$2]=$0;next} $1 in T {print T[$1], $0}' file2 file1
1       gi|358484498|ref|NW_003764396.1| gi|358484498|ref|NW_003764396.1| 2503 A G 0 1 1
1       gi|358484498|ref|NW_003764396.1| gi|358484498|ref|NW_003764396.1| 381 A G 1 1 2
1       gi|358484513|ref|NW_003764381.1| gi|358484513|ref|NW_003764381.1| 1351 T C 0 1 1

What exactly is wrong with this? It is the entire lines from file1 and file2 combined. It looks exactly like what you requested in post#1. What does "I do not get the same line in file1 which it had before" mean?

Dear Rudic

When I applied your command, I found 1613 common id between file1 and file2. However, file1 has originally 1805 id and file2 has originally 11580 id. because file1 is the master id file, after merging with file2 we expect 11580 lines which I got 1613 line by your command.

It prints only those lines in which $1 resp. $2 do match. If that's not the desired operation, please explain what this (from post#1) means:

Dear Rudic

As my first post, it is clear I want to join the common id between these two files (file1 and file2). But the point is file2 has id with replication and file1 just have unique number for each id. Usually after join two files, the unique ids should match with their ids which repeated in file1. In conclusion, I am trying to merge file1 to file2 without changing the lines of file2.

As in your 1st post; it is NOT clear.

And, I repeat: If you are unwilling to select a representative set of data from your input files that shows lines that should match, lines that should not match, and lines that have multiple matches; and show us the exact output you want from that sample input, then we are all wasting our time here.

Show us sample input. Show us the exact output that should be produced from that sample input If you are unwilling to do that, I will close this thread.