average by user defined cutoff

[kumarboston]

Hi all,
I was trying to calculate the average value from different parts of the same data file. For example, if suppose we have number 1 - 10 and i was trying to calculate the average of only first 3 values and then 4 values and then last 3 value and then calculate the three averages. I have written a code but I guess it is very good way to calculate it and sometimes i get garbage values also.

Here is a data file:
[CODE=perl]
1
2
3
4
5
6
7
8
9
10
[/CODE]

so when the user specifies the different parts as arguments, the average should be calculated.

Here is my perl code:
[CODE=perl]
#!/usrbin/perl

use strict;
use warnings;

my $file = $ARGV[0];
my $cut1 = $ARGV[1];
my $cut2 = $ARGV[2];
my $cut3 = $ARGV[3];
my (@tt,$result1,$ result2,$result 3);

open (A,$file);
my ($count,$total) = 0;
my ($val,$result);

while (<A>)
{
my @temp = split (/\s+/,$_);
$val = $temp[1];
$total = $total + $val;
$count++;

if ($count == $cut1)
{
$result1 = sprintf("%.3f", $total/$count);
push(@tt,$resul t1);
my $s1= $total;
my $r1 = $total/$count;
print "$s1\t$count\t$ r1\n";
}

if ($count == ($cut2+ $cut1))
{

$result2 = sprintf("%.3f", ($total-($result1*$cut1 ))/($count-$cut1));
push(@tt,$resul t2);
my $s2=($total -($result1*$cut1 ));
my $c2 = ($count - $cut1);
my $r2 = $s2/$c2;
print "$s2\t$c2\t$r2\ n";

}
if ($count == ($cut3+$cut2+$c ut1))
{
$result3 = $total- (($result1*$cut 1) + ($result2*$cut2 )) / ($count- ($cut1+$cut2));
push(@tt,$resul t3);
my $s3 = ($total- (($result1*$cut 1) + ($result2*$cut2 )));
my $c3 = $count - ($cut1+$cut2);
my $r3 = $s3/$c3;
print "$s3\t$c3\t$r3\ n";
}
}
my $add = 0;
foreach my $r(@tt)
{
$add = $add +$r;
}
print "$add/scalar(@tt)\t";
my $final = sprintf("%.3f", $add/scalar(@tt));
print "$final\n";
[/CODE]

Right now i can take 3 user cutoffs but if i want to make this program take any number of cutoffs to calculate the averages.
I guess there must a better way to calculate the average from different section of same data file.
Here I have used just 1 to 10 numbers as examples but my actual data files have 16900 lines and i have to calculate the average by using different parts of the file.
Any help in this regard is appreciated.
Thanks
Kumar

[KevinADC]

If you have a file with just numbers on each line, why are you using split?

 my @temp = split (/\s+/,$_);

Anyway, something like this seems easier:

#!/usr/bin/perl                                                                                                                                              
 
use strict;                                                                                                                                                 
use warnings;                                                                                                                                               
 
my $file = $ARGV[0];
my $cut1 = $ARGV[1];
my $cut2 = $ARGV[2];
my $cut3 = $ARGV[3];

my (@cut1, @cut2, @cut3);

open (my $IN, $file) or die "$!";
push @cut1, chomp <$IN> for (1..$cut1);
push @cut2, chomp <$IN> for ($cut1+1..$cut2);
push @cut3, chomp <$IN> for ($cut2+1..$cut3);
close $IN;

average(\@cut1,\@cut2,\@cut3);

sub average {
   my @arrays = @_;
   my $sum;
   foreach my $list (@arrays) {
      $sum += $_ for @{$list};
      my $avg = $sum / @{$list};
      print "Average = $avg\n";
      $sum = 0;
   }

[kumarboston]

Thanks KevinADC for the reply,
I ran your code on a file with numbers 1 to 10, but its throwing an error, "Can't modify <HANDLE> in chomp at new.pl line 14, near "<$IN> for "
Execution of new.pl aborted due to compilation errors."
I checked on the error and when I removed the chomp it was working fine but now "Illegal division by zero at new.pl line 27." error is coming and was not able to remove it.

Thanks
Kumar

[KevinADC]

oops, my bad. THis new version of the code assumes (for a 10 line file) that cut1 cut2 and cut 3 are equal to 3, 4, 3 respectively, if not the for() loop conditions need to be adjusted.

use strict;
use warnings;

my $file = $ARGV[0];
my $cut1 = $ARGV[1];
my $cut2 = $ARGV[2];
my $cut3 = $ARGV[3];
 
my (@cut1, @cut2, @cut3);
 
open (my $IN, $file) or die "$!";
for (1 .. $cut1) {
   chomp ($_ = <$IN>);
   push @cut1,$_;
}
for ($cut1+1 .. $cut1+$cut2){
   chomp ($_ = <$IN>);
   push @cut2, $_;
}
for ($cut1+$cut2+1 .. $cut1+$cut2+$cut3){
   chomp ($_ = <$IN>);
   push @cut3, $_;
}	
close $IN;
 
average(\@cut1,\@cut2,\@cut3);
 
sub average {
   my @arrays = @_;
   my $sum;
   foreach my $list (@arrays) {
      $sum += $_ for @{$list};
      my $avg = $sum / @{$list};
      print "Average = $avg\n";
      $sum = 0;
   }
}

[kumarboston]

Thanks so much Kevin,
the output for the calculation of 1 to 10 " Average = 2 ; Average = 5.5; Average = 9" is correct, only a last small help, if in the end if I want to take the average of these values then how do i modifiy the subroutine?

Thanks
Kumar

[KevinADC]

change the subroutine to something like this:

sub average {
   my @arrays = @_;
   my $sum;
   my @averages;
   my $total;
   foreach my $list (@arrays) {
      $sum += $_ for @{$list};
      push @averages, $sum;
      my $avg = $sum / @{$list};
      print "Average = $avg\n";
      $sum = 0;
   }
   ($total += $_) for @averages;
   printf "Total Average = %.3f\n", $total / @averages;
}

Next time, please show some effort on your part first to solve the problem.