How to select, copy and paste some columns from different files and create one file from them?

This can also be done quite easily with awk.

$ awk '{print $2,$3,$4}' *.txt > collapsed_output.txt

Example

Here's some sample data.

$ seq 20 | paste - - - - - > sample.txt

Here's what the lines look like:

$ head sample.txt 
1   2   3   4   5
6   7   8   9   10
11  12  13  14  15
16  17  18  19  20

Now let's make 10 copies:

$ seq 10 | xargs -I{} cp sample.txt sample{}.txt

We now have the following files:

$ tree
.
|-- sample10.txt
|-- sample1.txt
|-- sample2.txt
|-- sample3.txt
|-- sample4.txt
|-- sample5.txt
|-- sample6.txt
|-- sample7.txt
|-- sample8.txt
|-- sample9.txt
`-- sample.txt

Now if we run our awk command:

$ awk '{print $2, $3, $4}' sample{1..10}.txt | column -t
2   3   4
7   8   9
12  13  14
17  18  19
2   3   4
7   8   9
12  13  14
17  18  19
2   3   4
7   8   9
12  13  14
17  18  19
...

Here I'm showing you the output for the first 3 files (sample01.txt ... sample03.txt). Also I'm formatting the output with the column -t command, but this is only for display purposes to make the output easier to see here on U&L.

Additional formatting could just as easily been done within the awk command but that seemed to be beyond the scope of the question.

Have a look at the command line utility named cut. It can extract columns if they are separated by a unique delimiter. To recombine the parts you can use paste.

If you have, for example a typical comma-separated format

$ cat debts.csv
Name,Age,Debt
Alice,20,1337
Bob,30,42

$ cat pets.csv
Name,Pet
Alice,Dog
Bob,Cat

you could extract names and debts with

$ cut -d, -f1,3 debts.csv
Name,Debt
Alice,1337
Bob,42

and combine debts with pets using

$ cut -d, -f2 pets.csv | paste -d, debts.csv -
Name,Age,Debt,Pet
Alice,20,1337,Dog
Bob,30,42,Cat

• With cut and paste, -d determines the delimiter for the fields,
• -f selects the columns to extract for cut and
• - directs to use the standard input (i.e. in the latter paste case, from the pipe) instead of a file.