use Benchmark qw(timethese);
use strict;
use warnings;
my @array = (1 .. 100000);
timethese(100, {granfather => \&grandfather, usual => \&usual});
sub grandfather {
return 1 if grep {$_ == 100000} @array;
}
sub usual {
for my $a (@array) {
return 1 if ($a == 100000);
}
return 0;
}
If the element we serach is at the end, grandfather's is a little bit faster:
Benchmark: timing 100 iterations of granfather, usual...
granfather: 4 wallclock secs ( 3.58 usr + 0.00 sys = 3.58 CPU) @ 27
+.95/s (n=1
00)
usual: 4 wallclock secs ( 4.13 usr + 0.00 sys = 4.13 CPU) @ 24
+.24/s (n=1
00)
But if you change the above code to search for 3 (at the beginning of the array), then the usual way is much faster:
Benchmark: timing 100 iterations of granfather, usual...
granfather: 3 wallclock secs ( 3.58 usr + 0.00 sys = 3.58 CPU) @ 27
+.95/s (n=1
00)
usual: 0 wallclock secs ( 0.00 usr + 0.00 sys = 0.00 CPU)
(warning: too few iterations for a reliable count)
With something in the middle say 50000, the usual way is still close to 50% faster.
Benchmark: timing 100 iterations of granfather, usual...
granfather: 4 wallclock secs ( 3.69 usr + 0.03 sys = 3.72 CPU) @ 26
+.89/s (n=1
00)
usual: 2 wallclock secs ( 2.08 usr + 0.00 sys = 2.08 CPU) @ 48
+.12/s (n=1
00)
Now you know which direction it is pointing to ;-) |