#441: Run new performance profiler to see if there are places we can improve it

[cwschilly]

Fixes #441

[cwschilly]

While some of the "optimizations" seem trivial, others are good catches. For instance, the profiler caught this duplicated code in the run() function of LBAF_app.py:

for phase_id in self.__parameters.phase_ids:
    # Create a phase and populate it
    phase = Phase(
        self.__logger, phase_id, reader=reader)
    phase.populate_from_log(phase_id)
    phases[phase_id] = phase
    phase = Phase(
        self.__logger, phase_id, reader=reader)
    phase.populate_from_log(phase_id)
    phases[phase_id] = phase

---

I timed the execution of the run function through multiple iterations of adding these optimizations. Averaging the times from three runs on challenging-toy-hundreds-tasks.yaml, here are the results:

Change	Average run() time (s)
Original state	66.327
Change print_statistics call	52.208
Change arrangement sorting	59.030
Remove duplicated phase block	38.445
Rework qoi_request list	45.639

These are cumulative results.

---

Clearly some changes were more impactful than others. The timings also varied a lot, so it's not worth going by these alone. Let me know @ppebay @lifflander if you want to keep any/all of these changes. I'll update this table if I make any more revisions.

[cwschilly]

Thanks @ppebay , I made the requested change.