Reliance on private EGI chanlocs file leads to flatline channel label mismatch #17
Description
HAPPE v4 prompts users to enter a flatline channel (if one exists), identifying it by channel label. When analyzing data collected using 128-channel EGI Hydrocel nets, I entered 'E129' (no quotes), as this was the channel label for the system we used. However, it appeared that HAPPE was unable to match this up with an existing channel label when looking to remove the flatline channel.
Upon further investigation, I saw that HAPPE ignored the existing chanlocs, and instead relied on its own private chanlocs file for the appropriate EGI system. Its private chanlocs file uses the label 'Cz' for the reference channel; as a result, it was unable to find a match for 'E129'. (I can provide a sample data file and HAPPE parameter file if needed, but I expect this behavior is typical.)
I assume 'Cz' is used because it's standardized across EGI net sizes and/or because the name of the reference channel has historically varied even within a net size (e.g., the channel label for the 128-channel nets was 'E1001' on some older systems). However, it is not apparent to the user that, in response to the flatline prompt, they should use an alternate channel label which does not actually appear in their data file.
If the decision not to use the channel labels in the imported data file was a deliberate one, one possibility would be to check the user response to the flatline prompt and, if it matches a channel label in the imported data set, to convert that response to the channel label with the corresponding index in the private chanlocs file (e.g., 'E129' -> matches old index 129 -> convert to new index 129 -> 'Cz').