Closed neurofractal closed 4 years ago
Using Zapline with shorter windows does seem to help:
N.B. Please see compare PSD @ 50Hz between (left) applying Zapline to full dataset (800s) and (right) applying in 1s overlapping windows
Now implemented HERE
Zapline seems to work better for short 10s segments of data (probably lots of non-stationarities).
I should implement a sliding/overlapping window approach for our data