TVB-2371 Implement signal energy computation

[kimonoki]

The energy is filtered and computed based on the time window size in the time series datatype.
The result is calculated for every time point in the data. For the time window length that will exceed the last time point, the energy is computed to the last time point.
For example for time data [1,2,3] and length 2 the last energy will only calculate at time 3.

[maedoc]

The contribution is appreciated but implementation is not quite what I'd expect. It would be far more efficient to window the time series, apply a window function, FFT the windows, and sum over coefficients in the frequency band of interest. You might look at the coherence analyzer to get started.

Please avoid nested functions, and be sure to test your code, as I'm quite sure line 388 is wrong here

         "Length": self.sample_period * self.read_data_shaperead_data_shape('data')[0]}

[maedoc]

Overall this is a nice addition. Please make the filter frequencies parameters exposed to the user, and don't hard code the filter coefficients.

[maedoc]

This PR is getting stale without response to my review, and our development has moved to tvb-root repo. If you want to take this up again please send a fresh PR there.