trace2FWkurto calculates cumulative kurtosis over bandwidths and windows

Usage:
[mean_M,C]=trace2FWkurto(f,h,FB,T,n_smooth,first,last)

The FW in the name stands for Frequency Window because it computes for
several frequency bandwidth and window sizes

Input: ‘f’ is the raw trace (works on array, not on matrix)
‘h’ is the sampling frequency
‘FB’ is an n-by-2 matrix containing n frequency band
specifications
‘T’ is an m-by-1 vector of window lengths (seconds)
‘first’ is the first sample of interest
‘last’ is ,the last sample of interest
Ouput: ‘M’ is the (m*n)-by-(length(f)) kurtosis cumulative matrix
‘sum_trace’ is the mean value of the kurtosis for all windows
and all frequency bandwidth
Example: trace2FWkurto(f,100,[5 10;15 20],[1 2 3],10,1,1500)