v0.4

[lopezjuanma96]

Retake Time Feature Processing from v0.3:

• Check shape of array before preprocess
• If shape of array is not compatible with preprocess but the transposed is, ask if to transpose or not.
• FV generator in time series
• FV matrix in time series
• plot average on fv matrix in time
• How to save the FV matrix (use save json or create save numpy)
• Wavelet filtering.
• Wavelet Transform.
• FV Matrix From Time to Frequency

Extra features:

• Command Simulation in intels (even without the slider bars, maybe with the parameters as a function argument.
• Timestamps_loc for endpoints too

Add Simulated Signal Generator, this can have:

• Generate with sine waves samples.
• Generate with a sample signal or a sample numpy array.
• Properties: Channels, stimulation periods, resting periods, stimulation frequencies.
• For SSVEP and for P300.

[lopezjuanma96]

Modifications on the feature_vector_gen_neighbour function: this had a loop over the interest freqs that were not None, and on this loop created a mask that was then applied to bins and values of the FFT. But this created a mask that was interest_freqs number of times the shape it should be. Then a change was made to make the same loop but on each condition of the ifs statements. But then max_bins, that use to work on each loop for each interest freq, now will be for the global feature_vector, on all freqs, this then should be corrected or deleted to only use the bw and not the max_bins, value.