Block average of signal

Take the average of blocks in the signal. The best match using correlation can be used

USAGE
data_mean = block_average(data,start_end)
data_mean = block_average(data,start_end,shift)

INPUT
- DATA: NxM matrix with M signals along the rows
- START_END: Bx2 matrix with the first and last elements to take the average
- SHIFT: maximum number of data points to shift the elements in START_END to find the best match (greatest correlation coefficient). Only the first column in DATA is optimized. The same shift will be applied to all other columns.

OUTPUT
- DATA_MEAN: average of DATA along the rows for the elements in START_END

EXAMPLE
>> data_mean = block_average(data,[5 35;35 65;65 95],3);