[Discussion] Should NIDQ channels be treated like ElectricalSeries?

[h-mayorquin]

From a discussion with @bendichter about examples where we want to extend a column in the electrodes table.

While working on #985 I realize the that we are writing the nidq as an ElectricalSeries:

from neuroconv.converters import SpikeGLXConverterPipe

converter = SpikeGLXConverterPipe(folder_path=SPIKEGLX_PATH / "Noise4Sam_g0")
nwbfile = converter.create_nwbfile()


electrodes_df = nwbfile.electrodes.to_dataframe()
electrodes_df[electrodes_df["channel_name"].str[0] == "X"]

Reading the docs in the schema it seems that NIDQ does not really fit there:

https://github.com/NeurodataWithoutBorders/nwb-schema/blob/63ac845b2e1afdf6fc79466c1feafdd34538bb10/core/nwb.ecephys.yaml#L1-L7

Quoting:

A time series of acquired voltage data from extracellular recordings ....

(bold mine)

Maybe it is OK to write NIDQ as ElectricalSeries because the units are volts but what seems odd to me is that we are also writing the channels in the electrodes table.

I feel this should not be the case?
Thoughts?

[h-mayorquin]

OK, after discussing this at the team meeting the conclusion was that it should be a TimeSeries.