bug in pop_chanedit

[jiversen]

Description

pop_chanedit returns a malformed command to eegh. Running that command does not recreate the behaviour that was done in the gui.

Steps to Reproduce:

1. I started pop_chanedit from the menu on a 7-channel dataset that has only 1020 channel labels , but no channel locations or channel types. [The main window lists channel locations "No (labels only)"].

   I picked the standard head model then set all of the channel types to 'EEG'--this properly populates the channels with all of the standard coordinate values and returns a dataset with full chanlocs. Fine.

2. However, I then took the command string returned by eegh and added it to my own code

3. Running this command on a new dataset does not work: it returns an EEG.chanlocs with empty coordinates!

Expected behavior:

Running the code returned by pop_chanedit should properly lookup the electrode coordinates, and recreate the GUI behavior, not return empty coordinates.

Actual behavior:

The command string that was returned to eegh for the action was malformed and does not run properly:

EEG=pop_chanedit(EEG, {'lookup','/XXXX/eeglab/plugins/dipfit/standard_BEM/elec/standard_1005.elc'},'changefield',{1,'type','EEG'},'changefield',{2,'type','EEG'},'changefield',{3,'type','EEG'},'changefield',{4,'type','EEG'},'changefield',{5,'type','EEG'},'changefield',{6,'type','EEG'},'changefield',{7,'type','EEG'});

Notice how the second argument is a cell array -- this case is not handled in the pop_chanedit argument parsing code around line 248 (the case when orichaninfo is a cell is not handled, only the case when is a char).

This leads to the lookup param/value pair to be completly ignored, so the returned EEG.chanlocs has no coordinate values.

Suggested fix:

Fix the code that builds the command string. It should return:

EEG=pop_chanedit(EEG, 'lookup', '/XXXX/eeglab/plugins/dipfit/standard_BEM/elec/standard_1005.elc', 'changefield',{1,'type','EEG'},'changefield',{2,'type','EEG'},'changefield',{3,'type','EEG'},'changefield',{4,'type','EEG'},'changefield',{5,'type','EEG'},'changefield',{6,'type','EEG'},'changefield',{7,'type','EEG'});

Note there are no brackets around the lookup/path PV pair. This will be properly parsed by pop_chanedit

Workaround

Manually reformat the command you copied from eegh to remove those curly brackets.

Versions

OS version	macos 15.4.1
Matlab version	2024b
EEGLAB version	2025.0.0

[arnodelorme]

I am pretty sure this was fixed in 2025.1. Would you mind checking?

[jiversen]

Sorry, should have checked that first! Sure, I'll update.

[arnodelorme]

Actually maybe use the GitHub version because the problem might have been fixed after 2025.1 (I can’t remember). I confirmed that it is fixed for me right now.

…

On May 8, 2025, at 16:07, jiversen ***@***.***> wrote: jiversen left a comment (sccn/eeglab#867) Sorry, should have checked that first! Sure, I'll update. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: ***@***.***>