I am new to wavelets, but I would like to reproduce this processing step (from this paper) (p. 3 - Wavelet Decomposition paragraph):
Coiflet-1 is chosen as the wavelet basis [...] Raw EEG signal of 1-second epochs are normalized and the wavelet decomposition is performed at scale 2**3 (node 3 in wavelet package, corresponding to the frequency band 47−63Hz)
The coif1 wavelet has these properties:
But I don't know how it relates to the "scale" mentioned. Also, the signal used in the paper was sampled at 250Hz. Does that mean that we need to resample the signals to that particular sampling rate to get the same results? Or can we somehow adjust the scale / decomposition parameters to take into account the sampling rate?
Hi, and thanks for that package!
I am new to wavelets, but I would like to reproduce this processing step (from this paper) (p. 3 - Wavelet Decomposition paragraph):
The
coif1wavelet has these properties:But I don't know how it relates to the "scale" mentioned. Also, the signal used in the paper was sampled at 250Hz. Does that mean that we need to resample the signals to that particular sampling rate to get the same results? Or can we somehow adjust the scale / decomposition parameters to take into account the sampling rate?
Thanks for any pointers!