# -*- coding: utf-8 -*-

import numpy as np

import matplotlib.pyplot as plt

from pylab import *

# Normalisation function

# A is the matrix to normalise

# the treshold to remove bins for the normalization (these bins are replaced by zeros vectors)

def scn_func(A,threshold):

n1 = A.shape[0];

n_iterations=10;

keep = np.zeros((n1, 1));

for i in range(0,n1) :

if np.sum(A[i,]) > threshold:

keep[i] = 1

else :

keep[i] = 0

indices1=np.where(keep >0 )

indices2=np.where(keep <=0 )

for n in range(0,n_iterations) :

print(n);

for i in range(0,n1) :

A[indices1[0],i]=A[indices1[0],i]/ np.sum(A[indices1[0],i])

A[indices2[0],i]=0

A[np.isnan(A)] = 0.0

for i in range(0,n1) :

A[i,indices1[0]]=A[i,indices1[0]]/ np.sum(A[i,indices1[0]])

A[i,indices2[0]]=0

A[np.isnan(A)] = 0.0

return A