I was trying to get power spectral density of a time-series data. The sleep recording software we use, doesn't allow to export EDF format of an interval, only of the whole record, but it allows to extract an ASCII file with
uV value of a single lead. We are selecting the visually "cleaner" part of a longer dot test task.
The way I'm proceeding now, is to
- import them in eeglab,
- use FIR automatic filter with 0.5 - 28 low hi pass
run this script
[power, freq] = spectopo(EEG.data(chanNr, :), 0, sampRate, 'winsize', 512, 'freqfac', 2,'freqrange',[0 28]); (chanNr = 1 and sampRate =512)
I am not familiar with signal analysis, am I proceeding in the proper way? The output I get from
spectopo in the
power variable, is
uV squared, right?