I am trying to do imagined digit classification from EEG signals. I have collected EEG signals where I envision each of the digits 0 → 9. I have a big recording of size (n_samples, 14), its 14 channel EEG data. I want to compute band powers of EEG signal via the Morlet wavelet transform and transform this dataset into a dataset with dimensions (n_samples, 70). 70 because I want to compute band powers across five different frequency intervals, and there are 14 channels, 14 * 5 = 70. I am a beginner and I have no idea how to do this, can someone help me?
If you don’t want to epoch your data you should first start by reshaping in a compatible shape like data = data.reshape((1,14,n_samples)).
Then apply the transformation. I suggest you first take a look of some examples.
Finally, proceed depending on the parameters you used.
Note that, in any case, you are required to provide an array indicating which frequencies you are interested in. This means you will have one value per frequency bin of the transformation and not one value per transformation, unless your frequency intervals correspond to exactly one bin.