Conditional Bitwise XOR(^) Operator in Matching Chars Count of Two Strings?

monkfan has asked for the wisdom of the Perl Monks concerning the following question:

Guys,
I want to use bitwise XOR operator which is known to be fast for bitwise testing. What I intend to do below is to find number of matching positions (except N matching is not counted).

my $s1 = "STCATTNNNSATCGCT"; 
my $s2 = "ATCGTCGSNNNNATCG";
           ## #   0     #

# There are 4 matching positions - marked with (#) -
# and notice that matching N is not counted (0)
[download]

My best attempt with it is this, which leads me nowhere.

my $match = $s1 ^ $s2;
my $match_count = $match =~ tr/\00//;
print "$match_count\n";

# This gives 5, because N is counted.
# How can I make it to return 4 instead?
[download]

Because the code I have below - using subscripted for loop - already does its job, but it is horribly slow:


get_match_count($s1,$s2);

sub get_match_count {

    my ( $k, $l ) = @_;
    return 0 if $k eq $l;

    my $len       = length($k);
    my $num_match = 0;

    for ( my $i = 0; $i < $len; $i++ ) {
        next if ( substr( $k, $i, 1 ) eq 'N' 
                        and 
                  substr( $l, $i, 1 ) eq 'N' );
        ++$num_match if substr( $k, $i, 1 ) eq substr( $l, $i, 1 );
    }

    print "$num_match\n";
    return $num_match;
}
[download]

Regards,
Edward

Comment on Conditional Bitwise XOR(^) Operator in Matching Chars Count of Two Strings? Select or Download Code

Replies are listed 'Best First'.
Re: Conditional Bitwise XOR(^) Operator in Matching Chars Count of Two Strings by GrandFather (Saint) on Nov 09, 2005 at 02:04 UTC
Mutate the Ns first: `use strict; use warnings; use Data::Dumper; my $s1 = "STCATTNNNSATCGCT"; my $s2 = "ATCGTCGSNNNNATCG"; $s1 =~ tr/N/n/; my $merge = $s1 ^ $s2; my $count = $merge =~ tr/\0//; print $count;` [download] Prints: `4` [download] Perl is Huffman encoded by design.	[reply] [d/l] [select]
Re^2: Conditional Bitwise XOR(^) Operator in Matching Chars Count of Two Strings by inman (Curate) on Nov 09, 2005 at 11:41 UTC
Changing the transliteration line to `tr/SACTG/_____/c;` allows you to specify the characters that you want to keep. This maybe easier to work with in a biomed application.	[reply] [d/l]