I was testing the method for finding
peaks in grain boundary distributions.
It seem that it is quite efficient, but
some control parameters and threshold
must be adjusted. I have been playing with
differen values of those parameters
for various data sets.