IntanRecordingInterface should not be allowed to write streams other than the amplifier channels

[h-mayorquin]

From the current discussionin #789 it seems that IntanRecordingInterface should only write data with stream_id=0: the amplifier channels. Reason being that we write the data as an ElectricalSeries following the ecephys pipeline. This means that if any other stream is choosen the data will be incorrectly written.

The current signature of IntanRecordingInterface allows the user to specify the stream_id which leads to the problem described above.

neuroconv/src/neuroconv/datainterfaces/ecephys/intan/intandatainterface.py

Lines 77 to 95 in fa6edd5

	def __init__(
	self,
	file_path: FilePathType,
	stream_id: str = "0",
	verbose: bool = True,
	es_key: str = "ElectricalSeries",
	):
	"""
	Load and prepare raw data and corresponding metadata from the Intan format (.rhd or .rhs files).
	Parameters
	----------
	file_path : FilePathType
	Path to either a rhd or a rhs file
	stream_id : str, optional
	The stream of the data for spikeinterface, "0" by default.
	verbose : bool, default: True
	Verbose
	es_key : str, default: "ElectricalSeries"

Should we change this to avoid this scenario?
Pros and cons?

[CodyCBakerPhD]

If Intan as a format is that reliable in terms of a particular stream ID mapping onto proper neural channels, then sure we can just restrict the interface to that stream

The argument is usually left open for other formats though out of necessity to select which neural stream in the event of multiple probes, channel groups, bands, etc.

[h-mayorquin]

Yeah, stream_0 is fixed at neo to be the amplifier channels.

It is not clear to me if all the amplifier channels should always be ElectricalSeries but at that point there is nothing we do for now.