Tobias Pfeiffer

If anyone looks at this and says "SHOW ME THE CODE" - it's really not that much different, you can check https://github.com/bencheeorg/benchee/blob/master/lib/benchee/benchmark/runner.ex The relevant parts that take the run time...

I know I have my own level of fun talking to myself here, something that just came to my mind: With `inputs` the input/data is already "inside" the process we...

memory consumption is measured in the HTML samples I posted, and its exactly the same - was also my hope that this would show something. Maybe amount of GCs could...

:wave: Thanks for the report! Also, I wouldn't trust whatever output you received. Once it starts swapping performance will be abysmal where it swapped. I'm surprised that it generated so...

@goncalotomas that the time explosion comes after the benchmarks is interesting behavior, maybe I misread the initial post :thinking: That leads to thinking more that it's the statistics/sorting part. In...

Ugh good point about us doing that in parallel :D that means there are 100 sorts running in parallel here which really explains this. Imo both 1 and 2 would...

Ah btw. great thought of fixing that through offering 2. @devonestes - never occurred to me but seems great :rocket: :star: :green_heart:

I haven't looked at it from the perspective of O-notations. That might be doable though. I've looked at it more from the perspective of: * finding performance edge cases. Generators...

also in case anything hits a cache on any layer (even cpu lvl...) then more varied inputs assure more accurate results :)

Yeah I was thinking in both ways :) And I'm not quite sure which one I like better. E.g. cases: 1. run each input a certain amount of time (gives...