For fun rather than for serious purpose, using the reduce function of the List::Utils module (or a home-made reduce function*) to build a library of statistical utilities in functional programming style.

sub sum {
    reduce { $a + $b } 0, @_; 
}
sub avg {     
    return undef unless @_; # avoid division by 0
    sum (@_)/scalar @_;
}
sub variance {
    return undef unless @_; # avoid division by 0
    my $avg = avg (@_);
    sum ( map {($_ - $avg)**2} @_ )/ scalar @_;
}
sub std_dev {   
    sqrt variance (@_);
}
[download]

The reduce utility can also be written in pure Perl:

sub reduce (&@) {
    my $code_ref = shift;
    my $result = shift;
    for (@_ ) {
        local ($a, $b) = ($result, $_ );
        $result = $code_ref->($a, $b )
    }
    return $result;
}
[download]