Hanno Spreeuw

Please assign this to me for the next sprint.

We just had a meeting with @sverhoeven and Adam. @sverhoeven had some clear ideas about how to proceed. Obviously we'll create a profiling branch and folder. Separate out the `test_user_workflow`...

I profiled [the modified code from Florian](https://github.com/matchms/matchms/issues/238#issuecomment-630791026), after making a small modification to the lower limit: `s = select_by_relative_intensity(s, intensity_from=0.5, intensity_to=1.0)` and the input file, this is now the 60...

@sverhoeven : https://www.pluralsight.com/guides/profiling-memory-usage-in-python Mentions memory-profiler (in pypi) and line_profiler (apparently not in pypi).

Bouwe said we should also check the guide: https://guide.esciencecenter.nl/#/best_practices/language_guides/python?id=profiling

Output from `python -m memory_profiler profiling.py` on the `profiling` branch: ``` Filename: /matchms/matchms/Spikes.py Line # Mem usage Increment Line Contents ================================================ 8 500.062 MiB 500.062 MiB @profile 9 def __init__(self,...

No, not intentional. How did you derive that number?

I removed the `5*` from the profiling.py script (twice) and all the `@profile` decorators in that script and the matchms code - which was required for the memory-profiler - and...

Done within a few minutes. I can easily inspect cachegrind.out using kcachegrind.

Perhaps adding the runtime for a plain run is also useful: ``` time python profiling.py real 0m27,872s user 0m26,437s sys 0m0,133s ```