[Eeglablist] averaging time-frequency maps in eeglab

[Kerick, Scott (Civ, ARL/HRED)]

Glad you posted this question as I was just about to. I have been using
the timef.m function in a script that loops through 32 chans x 36 trials
x 11 conditions for each of 18 subj. It's taking about 3 hours to run
one subj in one condition (1.8 GHz processor) so the overall computing
time will exceed 600 hrs! I think the process could be accelerated
significantly if plotting could be suppressed (opt noplots) but I
haven't found that to be an option. I've explored several other
alternatives external to EEGLAB (specgram, pmtm, dpss, pwelch, fft,
etc.) but would prefer to use timef because of its many added features
(phasecoher, etc.). Any suggestions from anyone would be appreciated.

Thanks,
Scott Kerick