 |
 |
Dear Duncan, Yes - the Morlet wavelet factor is indeed the number of cycles. I don't think decreasing it from 7 to a lower number will help you to estimate the lower frequencies. The fact is an 8Hz signal has 125ms period, so only 1.6 cycles of it are available in the 200ms baseline period (specifying a factor of 7 or 5 or 4 won't matter in this instance as the relevant wavelet basis function will be cut-off anyway). The only thing that will allow you to better estimate your 8Hz activity is to increase the baseline window length. But if you're already getting decent results, then it seems 1.6 cycles is already sufficent. Best, Will. Duncan Astle wrote: > Hi all, > > So, I am performing a time-frequency analysis on some EEG data. The data > are organised into -200 to +800 ms epochs, and I am interested in > frequencies between 8 and 40 ish Hz. When asked which "Morlet Wavelet > Factor" I want, I have left it on the default (7). The results look very > sensible indeed, but now I am worried that I shouldn't have used the > value '7'. > > Does the value asked for correspond to the number of cycles? In which > case surely I should have gone for a much lower value (otherwise, for > example, I won't be able to properly estimate the time-frequency > decomposition during the 200 ms baseline period)? > > But then again, the results look very sensible, so maybe I should just > stick with 7........ > > Any advice would be gratefully received. > > Cheers, > Duncan > > > -- William D. Penny Wellcome Trust Centre for Neuroimaging University College London 12 Queen Square London WC1N 3BG Tel: 020 7833 7475 FAX: 020 7813 1420 Email: [log in to unmask] URL: http://www.fil.ion.ucl.ac.uk/~wpenny/