Benchmark tool for benchmarking how long a block of code runs for X iterations or how many times it can run in a given period.
Bench now uses Telemetry if you have it to show you the USR/SYS/CPU times, so your output below may look slightly different.
use Bench;
my $b = Bench.new;
$b.timethese(1000, {
first => sub { sleep .05; },
second => sub { sleep .005; },
});
'---------------------------------------------------------'.say;
$b.cmpthese(1000, {
first => sub { sleep .05; },
second => sub { sleep .005; },
});
Output
Benchmark:
Timing 1000 iterations of first, second...
first: 51.5808 wallclock secs @ 19.3871/s (n=1000)
second: 6.4035 wallclock secs @ 156.1656/s (n=1000)
---------------------------------------------------------
Benchmark:
Timing 1000 iterations of first, second...
first: 51.5511 wallclock secs @ 19.3982/s (n=1000)
second: 6.4145 wallclock secs @ 155.8971/s (n=1000)
O--------O--------O-------O--------O
| | Rate | first | second |
O========O========O=======O========O
| first | 19.4/s | -- | -88% |
| second | 156/s | 704% | -- |
------------------------------------
Takes an array returned from any of the following methods and returns a formatted string with the data filled in. The string is similar to below: 6.4145 wallclock secs @ 155.8971/s (n=1000)
Times a single sub over X iterations. Doesn't output anything by default, just returns an array of time spent and iterations. Use in conjunction with .timestr
Returns how many iterations of the Sub it can run in the specified time. Use in conjunction with .timestr
Runs the specified sub for $iterations and automatically prints out the .timestr
. If $iterations is negative or 0 then it runs .countit
instead of .timeit
Similar to .timethis
but the key in the hash becomes the title for the test. An example of the output can be
Similar to .timethese
but it produces a cute little table comparing the results.
Do whatever you want with it.