Script for replacing columns

[Tlaloc86]

Hi, I am new to Linux, so I have a basic question. I want to convert a file like this:

ATOM 1 O SOL 2 20.662 13.241 18.640 1.00
ATOM 2 H SOL 2 21.639 13.282 18.849 1.00
ATOM 3 C SOL 2 20.497 13.637 17.737 1.00
ATOM 4 O SOL 3 0.535 11.254 21.810 1.00
ATOM 5 H SOL 3 1.097 12.079 21.746 1.00
ATOM 6 C SOL 3 -0.370 11.436 21.426 1.00

Into this:

ATOM 1 O SOL 2 20.662 13.241 18.640 4
ATOM 2 H SOL 2 21.639 13.282 18.849 2
ATOM 3 C SOL 2 20.497 13.637 17.737 3
ATOM 4 O SOL 3 0.535 11.254 21.810 4
ATOM 5 H SOL 3 1.097 12.079 21.746 2
ATOM 6 C SOL 3 -0.370 11.436 21.426 3

So basically, if the value at column "3" is O, H, or C, then I want to replace column "9" with 4, 2, or 3, respectively. I know I have to use the gawk or awk command.


Thank you for your help!

[pkleiber]

Hi, I am new to Linux, so I have a basic question. I want to convert a file like this:

ATOM 1 O SOL 2 20.662 13.241 18.640 1.00
ATOM 2 H SOL 2 21.639 13.282 18.849 1.00
ATOM 3 C SOL 2 20.497 13.637 17.737 1.00
ATOM 4 O SOL 3 0.535 11.254 21.810 1.00
ATOM 5 H SOL 3 1.097 12.079 21.746 1.00
ATOM 6 C SOL 3 -0.370 11.436 21.426 1.00

Into this:

ATOM 1 O SOL 2 20.662 13.241 18.640 4
ATOM 2 H SOL 2 21.639 13.282 18.849 2
ATOM 3 C SOL 2 20.497 13.637 17.737 3
ATOM 4 O SOL 3 0.535 11.254 21.810 4
ATOM 5 H SOL 3 1.097 12.079 21.746 2
ATOM 6 C SOL 3 -0.370 11.436 21.426 3

So basically, if the value at column "3" is O, H, or C, then I want to replace column "9" with 4, 2, or 3, respectively. I know I have to use the gawk or awk command.


Thank you for your help!

This would be more straightforward with R, but with auk this should do it:

[]$ awk '($3=="O"){x=4}($3=="H"){x=2}($3=="C"){x=3}{for(i= 1;i<=8;i++) printf "%s ", $i; print x;next}; {print $3,"not found"}' oldfile >newfile

Note the sanity check for bum records.

[beparas]

Thank you pkleiber
This is really good information.

[alf55]

Hi, I am new to Linux, so I have a basic question. I want to convert a file like this:

ATOM 1 O SOL 2 20.662 13.241 18.640 1.00
ATOM 2 H SOL 2 21.639 13.282 18.849 1.00
ATOM 3 C SOL 2 20.497 13.637 17.737 1.00
ATOM 4 O SOL 3 0.535 11.254 21.810 1.00
ATOM 5 H SOL 3 1.097 12.079 21.746 1.00
ATOM 6 C SOL 3 -0.370 11.436 21.426 1.00

Into this:

ATOM 1 O SOL 2 20.662 13.241 18.640 4
ATOM 2 H SOL 2 21.639 13.282 18.849 2
ATOM 3 C SOL 2 20.497 13.637 17.737 3
ATOM 4 O SOL 3 0.535 11.254 21.810 4
ATOM 5 H SOL 3 1.097 12.079 21.746 2
ATOM 6 C SOL 3 -0.370 11.436 21.426 3

So basically, if the value at column "3" is O, H, or C, then I want to replace column "9" with 4, 2, or 3, respectively. I know I have to use the gawk or awk command.


Thank you for your help!

Here is a "filter" written only in bash that will do what you want.

#!/bin/bash
#

while read line; do
    field=( ${line} )
    case "${field[2]}" in
        O)  field[8]=4 ;;
        H)  field[8]=2 ;;
        C)  field[8]=3 ;;
    esac
    echo "${field[@]}"
done