How to only store certain column from input file? - perl

Question

I am creating a script that should take in a data file and a column index, read and store that column and then perform some statistics on the data. I am unsure how to specify that I only want to store a specific column in Perl. Here is my code so far:
#! /usr/bin/perl
use warnings;
use strict;
use feature qw(say);
use Scalar::Util qw(looks_like_number);
my ($FILE, $COLUMN_TO_PARSE) = #ARGV;
#check if file arg is present
if(not defined $FILE){
die "Please specify file input $!";
}
#check if column arg is present
if(not defined $COLUMN_TO_PARSE){
die "Please specify column number $!";
}
unless(open(INPUT_FILE, "<", $FILE)){
die "Couldn't open ", $FILE ," for reading!", $!;
}
my #data;
while(<INPUT_FILE>){
# Only store $COLUMN_TO_PARSE, save to #data
}
close(INPUT_FILE);
For reference, the data coming in looks something like this(sorry for format):
01 8 0 35 0.64 22
02 8 0 37 0.68 9
03 8 0 49 0.68 49
So for example, if I ran
perl descriptiveStatistics.pl dataFile.txt 3
I would expect to have [35,37,49] in the #data array.
I stumbled upon this question, but it has to do with headers which I don't have, and not very helpful imo. Any suggestions?

Answer 1

I've used split() to split the input into a list of records. By default, split() works on $_ and splits on white space - which is exactly what we want here.
I've then used a list slice to get the column that you want, and pushed that onto your array.
#! /usr/bin/perl
use warnings;
use strict;
# Check parameters
#ARGV == 2 or die "Please specify input file and column number\n";
my ($file, $column_to_parse) = #ARGV;
open my $in_fh, '<', $file
or die "Couldn't open $file for reading: $!";
my #data;
while (<$in_fh>){
push #data, (split)[$column_to_parse];
}
If I was writing it for myself, I think I would replace the while loop with a map.
my #data = map { (split)[$column_to_parse] } <$in_fh>;
Update: To ensure that you have been given a valid column number (and I think that's a good idea) you might write something like this:
while (<$in_fh>){
my #fields = split;
die "Not enough columns in row $.\n" if $#fields < $column_to_parse;
push #data, $fields[$column_to_parse];
}

Answer 2

split is a good choice:
while (my $line = <INPUT_FILE>) {
my #items = split(/\t/, $line);
push #data,$items[$COLUMN_TO_PARSE];
}

Answer 3

You could design a regexp pattern for matching a column, which you repeat $COLUMN_TO_PARSE times, and which then captures the content of the column and pushes it onto your array #data.
Like this:
#!/usr/bin/perl
use strict;
use warnings;
use Data::Dumper;
my #data;
my $COLUMN_TO_PARSE = 3;
while (<DATA>) {
if (/([^\s]+\s+){$COLUMN_TO_PARSE}([^\s]+)/) {
push #data, $2;
} else {
print("error wrong line format: $_\n");
}
}
print Dumper(#data);
__DATA__
01 8 0 35 0.64 22
02 8 0 37 0.68 9
03 8 0 49 0.68 49
which gives the following dump for #data:
$VAR1 = '35';
$VAR2 = '37';
$VAR3 = '49';
$COLUMN_TO_PARSE is zero based, as in your example, and, as a side-effect, the regexp will fail if the requested column does not exist, thus giving you error-handling.

Answer 4

You can use split to get data column wise. Each column in stored in consecutive indices of the array.
while(<INPUT_FILE>){
my #columns = split(/\t/, $_); #Assuming delimiter to tab
print "First column====$columns[0]\n";
print "Second column====$columns[1]\n";
}
Do process whichever column you want and store into an array.