refactored and optimized segmentation metrics eval

[Mintas]

Refactored and optimized segmentation metrics evaluation.

1. Refactoring:
   key purpose is to make some of the methods more useful for extension in client libraries, make code more concise, reducing code duplications, thus also increasing possible community support

2. Optimisation:
   here we note, that some code duplications lead us to recalculate some preparative steps for metric calculation. Reusing common logical steps helped to significantly increase performance. In our microbenchmark of segment.eval() we have achieved 1.58 speedup!
   This may be crucial in some applications that may require intensive metric calculation (like Hyperparameter Optimization)

3. Benchmarking details:
   We took some symbolic music files (midis) with known expert markup of musical segments. There were 20 data points considered. We ran 300 iterations of eval method over all data provided to get better averaging.

Benchmarking results:
Before: time: 24.702 and latency: 0.0823, avg per data point: 0.0041
After: time: 15.615 and latency: 0.0520, avg per data point: 0.0026
Speedup is: 1.582 (~ -37%)

[Mintas]

@bmcfee @craffel
This PR is getting stale, would you consider reviewing and merging it?
TY

[bmcfee]

Sorry I won't have time to review this for a few weeks.

[bmcfee]

Sorry for the delay, I'm looking at this now. It's not the easiest PR to review because there's a mixture of cosmetic changes (ie variable names) and refactoring to sift through. It also appears that tests are not passing.

[bmcfee]

This would probably be cleaner using the decorator module, as we do in util, eg

mir_eval/mir_eval/util.py

Lines 946 to 962 in 935451d

	def deprecated(*, version, version_removed):
	"""Mark a function as deprecated.
	Using the decorated (old) function will result in a warning.
	"""
	def __wrapper(func, *args, **kwargs):
	"""Warn the user, and then proceed."""
	warnings.warn(
	f"{func.__module__}.{func.__name__}\n\tDeprecated as of mir_eval version {version}."
	f"\n\tIt will be removed in mir_eval version {version_removed}.",
	category=FutureWarning,
	stacklevel=3, # Would be 2, but the decorator adds a level
	)
	return func(*args, **kwargs)
	return decorator(__wrapper)

[Mintas]

Good note, gotta update my solution!

[bmcfee]

Why is this deprecated?

[Mintas]

It looks like this parameter is unused in this method implementation, can’t find good reason to continue its support.

[bmcfee]

Hrm, indeed. I wonder how this ended up there in the first place?

I agree that it makes sense to deprecate this. However, I think we should take that to a separate PR, and add in a bit of machinery to support deprecations better. (See, e.g., https://github.com/librosa/librosa/blob/ebd878fabb87c588f98f372a94fca5d167c59cf2/librosa/util/deprecation.py#L10-L15 )

[bmcfee]

I don't understand what these functions do - it seems like abstraction that may not be necessary, and it definitely does not help code readability.

[bmcfee]

This is a little unclear, since it's by default going to give trim is None rather than trim is False. The behavior will be the same (at least until we add type annotations), but it would be better to explicitly populate a boolean.