Task data missing expected fields

[neil-gallagher]

Processing and saving task data HAPPE produces an 'EEG' structure missing the 'epoch' and 'event' fields. I believe these are expected for EEGLAB-style data that has been split into epochs (described here). As a result, attempting to import a .set file of task data processed by HAPPE into MNE-python using read_epochs_eeglab results in an error.

[PINE-Lab]

Hi Neil,

To best be able to help you, we need some more information. What paradigm (rest/task) are you running? What file format is your original data in? Which acquisition layout/net/cap are you using? What were the input parameters that you selected?

Additionally, Please include the following information so we can best help you:

• Whether you are running on a Mac, PC, Linux, etc.
• MATLAB version
• HAPPE version (download date helpful but not necessary)
• Both quality assessment outputs, if generated (Pipeline QC and Data QC)
• HAPPE’s output Error Log (if created)

If you are willing and able to share a sample file with us, we can try to replicate the error on our end. However, we understand that sharing files is not always feasible. We are happy to meet via video conference to troubleshoot as necessary!

[neil-gallagher]

inputParameters_21-04-2023.mat.zip
I've attached the input parameters file that was output by HAPPE. We are using a 256 channel Hydrocel GSN net from EGI. The data is originally output as a .mff file. We are doing recordings during TMS stimulation, which requires some light preprocessing, which we do in EEGLAB and save this into a .set file before running HAPPE. I am running HAPPE 3.3 in MATLAB 2021b on an M1 macbook.

I was under the impression that not saving the 'epoch' and 'event' fields was a missing feature, rather than an error. Would you expect those fields to be saved in other segmented data?

[PINE-Lab]

Thanks for sending that over! Can you confirm that the 'epoch' and 'event' fields are present in the data that you feed into HAPPE? In other words, do your lightly preprocessed, converted to .set files have these fields before running HAPPE?

We expect those fields to be present, especially because HAPPE runs on data in EEGLAB format.

[neil-gallagher]

The 'event' field does exist in the .set files, but don't contain the 'epoch' field since they are still in the continuous time series form and I am using HAPPE to do the epoching.