I have a EEG time series data in a single vector form. The dataset consists of 32-seconds (approximately 16,400 samples) EEG signals recording with 256 Hz sampling frequency.
Can anyone please suggest me how the large 32-seconds EEG recordings can be segmented into an 8-seconds window size of 4 equal segments of 2048 samples?
Dear Muhammad Tariq Sadiq,
I suggest you to follow: https://www.mathworks.com/matlabcentral/answers/266255-segmenting-sinusoidal-time-series-data
https://www.mathworks.com/matlabcentral/answers/370337-data-segmentation-for-accelerometer-time-series-data
https://www.mathworks.com/help/ident/ref/segment.html
Non-overlapping windows or windows with partial overlap? For non-overlapping windows, s1=v(1:n/4), s2=v(n/4+1:n/2), s3=v(n/2+1:3n/4), s4=v(3n/4+1:n) , where v is the original vector with total n samples.
For the second case how much is the overlap needs to be known.
- Badges
- Science topic