So I ran a JMH Benchmark over a few methods and got response like this:
I am not able to understand what exactly the Score and Error
values signify.
Is there some reference documentation available for the same?
So I ran a JMH Benchmark over a few methods and got response like this:
I am not able to understand what exactly the Score and Error
values signify.
Is there some reference documentation available for the same?
As an example, ss_stream
with parameters n=100000
took ca. 1,363 microseconds to run on average over 30 iterations (and was run a number of times in each of these iterations, depending on the length of each iteration).
Assuming that the results are normally distributed, for that sample size, one would expect that the "true" execution time for that method has a 99.9% probability to be somewhere between 1,362.752 - 126.340 microseconds and 1,362.752 + 126.340 microseconds, i.e. between ca. 1,236 and 1,489.
As far as I know, the main documentation is in the samples. The second one gives information about the various benchmark modes.
© 2022 - 2024 — McMap. All rights reserved.