bash scripting: loop over a file, replacing two decimal numbers

frankie_DJ · 04-30-2007, 09:37 AM

Hi everyone,

I've been struggling to get the following done in a bash script:

I have a file with some data, and I want to replace two lines in that file so that it can be input to another program. But this replacement needs to be done 20 times, so 20 different replacements.

The lines are:

radius 1 0.57 1.40
...
...
radius 39 1.10 2.10

The two decimal numbers (in the last two columns) should be multiplied by the number x, which takes values
1, 1.05, 1.10,..., 1.50

The looping part is easy, but the actual replacement is what I'm having problem with. I know I should use sed or awk and bc. But how?

Thanks for your help.

theNbomr · 04-30-2007, 01:11 PM

Bash & sed are probably not up to the task, since neither of them is capable of floating point arithmetic. Awk or perl would probably be suitable. Your description of the requirement is still a little vague but this snippet of perl might get you somewhere close.

Code:

#! /usr/bin/perl -w

use strict;

 open( INFILE, $ARGV[0] ) or die "Cannot open $ARGV[0] : $!\n";
 my @inFile = <INFILE>;
 close INFILE;
 for( my $i = 1; $i <= 20; $i++ ){
  my @outFile = ();
  my $factor = 1 + ( .05 * $i );
  foreach my $inLine ( @inFile ){
    if( $inLine =~ m/radius\s+1\s+(0\.57)\s+(1\.40)/ ){
      my $col3 = $1 * $factor;
      my $col4 = $2 * $factor;
      push @outFile, "radius 1 $col3 $col4\n";
    }
    elsif( $inLine =~ m/radius\s+39\s+(1\.10)\s+(2\.10)/ ){
      my $col3 = $1 * $factor;
      my $col4 = $2 * $factor;
      push @outFile, "radius 39 $col3 $col4\n";
    }
    else{
      push @outFile, $inLine;
    }
  }
  open( OUTFILE, ">$ARGV[0]_$i" ) or die "Cannot write file '$ARGV[0]_$i' : $!\n";
  print OUTFILE @outFile;
  close OUTFILE;
 }

It should read the file, and create 20 new files subscripted with the index number, and with the numbers adjusted per the index.
--- rod.

cfaj · 04-30-2007, 05:04 PM

Quote:

Originally Posted by frankie_DJ

Hi everyone,

I've been struggling to get the following done in a bash script:

I have a file with some data, and I want to replace two lines in that file so that it can be input to another program. But this replacement needs to be done 20 times, so 20 different replacements.

The lines are:

radius 1 0.57 1.40
...
...
radius 39 1.10 2.10

The two decimal numbers (in the last two columns) should be multiplied by the number x, which takes values
1, 1.05, 1.10,..., 1.50

The looping part is easy, but the actual replacement is what I'm having problem with. I know I should use sed or awk and bc. But how?

Use awk:

Code:

{
  $2 *= nextval
  $3 *= nextval
}