Understanding `win_sec` in `bandpower` about yasa

[raphaelvallat]

Hi @matiasandina,

Your method should work well. Perhaps a more optimized method (though less intuitive) is to use:

1. yasa.sliding_window to convert your data into a 2D (or 3D) array of epochs (e.g. 4-sec epochs of data).
2. Then manually calculate the power spectrum on all epochs at once using scipy.signal.welch.
3. Finally, use yasa.bandpower_from_psd_ndarray to calculate the bandpower on all epochs at once. Note that the output will be a numpy array and not a pandas dataframe.

Alternatively, you could create a vector mask of the same length of the data where [ 1 1 1 1 ... 2 2 2 2 .... 3 3 3 3 ... ] 1 = first epoch, 2 = second epoch, etc. Then you can do: yasa.bandpower(data, sf, hypno=mask, include=(1, 2, 3)) which should calculate the bandpower for each value in mask separately.

Hope this helps,
Raphael

[matiasandina]

Perfect!