would it be faster to use an array?

If you mean "faster to use an array instead of a hash for collecting the data", then no, it would not be faster. I would split each line into an array during processing though.

The idea is to key the hash by the filenames. So, everytime you come across, for instance, "/some/dir/file.html", you increase a count and a sum. The code might look something like this (untested):

while (<LOG>) {
    my @part = (split ' ', $_)[5,8]; 
    $hash{$part[0]}->[0] ++;             # increase the count. 
    $hash{$part[0]}->[1] += $part[1];    # increase the sum.
}
[download]

 
for my $key (sort keys %hash) {
    my $avg = $hash{$key}->[1] / $hash{$key}->[0];    # sum/count.
    print "$key\t$avg\n";  
}
[download]

-sauoq
"My two cents aren't worth a dime.";