If #165 works, then the gridsearch does not need to be run to the hexels on the media wall.
In particular:
remove the fsaverage medial wall hexels from the search after averating and before the gridsearch, and then add then channels back in, with p-value=1 to them for all latencies after the gridsearch is completed.
make sure that the #SBATCH --mem=240G memory usage drops from 240GB down to a more suitable number in order that more runs can be on the queue nodes at once (currently 3 only allowed due to per user memory usage limits)
If #165 works, then the gridsearch does not need to be run to the hexels on the media wall.
In particular:
p-value=1
to them for all latencies after the gridsearch is completed.#SBATCH --mem=240G
memory usage drops from 240GB down to a more suitable number in order that more runs can be on the queue nodes at once (currently 3 only allowed due to per user memory usage limits)