Add get_times() method to imaging classes

[h-mayorquin]

We have set_times:

roiextractors/src/roiextractors/imagingextractor.py

Lines 189 to 212 in e8eb2cf

	def set_times(self, times: ArrayType) -> None:
	"""Set the recording times (in seconds) for each frame.
	Parameters
	----------
	times: array-like
	The times in seconds for each frame.
	Raises
	------
	ValueError
	If the length of 'times' does not match the number of frames.
	"""
	num_frames = self.get_num_frames()
	num_timestamps = len(times)
	if num_timestamps != num_frames:
	raise ValueError(
	f"Mismatch between the number of frames and timestamps: "
	f"{num_frames} frames, but {num_timestamps} timestamps provided. "
	"Ensure the length of 'times' matches the number of frames."
	)
	self._times = np.array(times).astype("float64", copy=False)

In neuroconv we use this construct to get the times:
https://github.com/catalystneuro/neuroconv/blob/419ab11104ff9e67d63f6400aacb3e12081c90d1/src/neuroconv/tools/roiextractors/roiextractors.py#L534-L535

Maybe we should have a method that does this for us like in spikeinterface?

https://github.com/SpikeInterface/spikeinterface/blob/bfe9fb649b58c48d2a04949061154f75f8e08b1f/src/spikeinterface/core/baserecording.py#L429-L450

Thoughts?

[pauladkisson]

We already have frame_to_time, which basically performs this function. But maybe it should be renamed? And/or maybe it should provide all the times by default?

[h-mayorquin]

Let's not have function with the same name than the spikeinterface one but with a different signature and behavior. The second option would work for me though.

I would prefer to unify the extractor API and just follow spikeinterface here. get_times can be a wrapper around frame_to_time if code duplication is your concern.