Monopolar triangulation for compute('unit_locations') is inconsistent with compute('spike_locations')

[ecobost]

compute_monopolar_triangulation() from localization_tools:

spikeinterface/src/spikeinterface/postprocessing/localization_tools.py

Lines 18 to 27 in 6b47603

	def compute_monopolar_triangulation(
	sorting_analyzer_or_templates: SortingAnalyzer | Templates,
	unit_ids=None,
	optimizer: str = "least_square",
	radius_um: float = 75,
	max_distance_um: float = 1000,
	return_alpha: bool = False,
	enforce_decrease: bool = False,
	feature: str = "ptp",
	) -> np.ndarray:

is used for analyzer.compute('unit_locations') while LocalizeMonopolarTriangulation from sortincomponents.peak_localization

spikeinterface/src/spikeinterface/sortingcomponents/peak_localization/monopolar.py

Lines 45 to 55 in 6b47603

	def __init__(
	self,
	recording,
	parents,
	return_output=True,
	radius_um=75.0,
	max_distance_um=150.0,
	optimizer="minimize_with_log_penality",
	enforce_decrease=True,
	feature="ptp",
	):

is used for motion correction and analyzer.compute('spike_locations').

The issue is that they have different default arguments (as can be seen above). Is that intended? I would think the args from LocalizeMonopolarTriangulation (which use more bells and whistles) make more sense overall.

A separate issue is that enforce_decrease has no effect in compute_monopolar_triangulation (it is just commented out)

spikeinterface/src/spikeinterface/postprocessing/localization_tools.py

Lines 118 to 121 in 6b47603

	# if enforce_decrease:
	# enforce_decrease_shells_data(
	# wf_data, best_channels[unit_id], enforce_decrease_radial_parents, in_place=True
	# )

Is that also intended? Either it should be uncommented or the argument dropped.