Python: distribution of shortest paths is irregular with regular grid

Question

In a regular NxN network, I want to color code the nodes based on the number of shortest paths passing through them. This is known in literature as the Stress Centrality (SC).

To do this, I use the nx.all_shortest_paths() function, which computes all the shortest paths between any two nodes in the graph.

The network in question is highly regular, so I firmly believe the distribution of shortest paths should follow the same pattern regardless of the size of the network.

But here is the deal: if the size is 9x9, it is clear the central nodes are the most "stressed", as seen below (nodes in white); if the size is 10x10, this cloud of stressed nodes moves elsewhere. I don't know if this is a Python/computational effect or if it is normal. I haven't tested networks bigger than 10x10 as it takes ages to do the calculations (time complexity seems to be exponential for this computation).

How can this happen? I expect the most stressed nodes to always stay in the center. Why is this not true when I increase the network size? After all, the topology remains unaltered (thus, symmetrical).

---

Images:

Code:

from __future__ import print_function, division
import numpy
import networkx as nx
import matplotlib.pyplot as plt
import matplotlib.colors as mcolors

#Creating the network
N=9
G=nx.grid_2d_graph(N,N)
pos = dict( (n, n) for n in G.nodes() )
labels = dict( ((i,j), i + (N-1-j) * N ) for i, j in G.nodes() )
nx.relabel_nodes(G,labels,False)
inds=labels.keys()
vals=labels.values()
inds.sort()
vals.sort()
pos2=dict(zip(vals,inds))
nx.draw_networkx(G, pos=pos2, with_labels=True, node_size = 15)

#Function counting all shortest paths between any two nodes
counts={}
for n in G.nodes(): counts[n]=0
for n in G.nodes():
    for j in G.nodes():
        if (n!=j):
            gener=nx.all_shortest_paths(G,source=n,target=j)
            for p in gener: 
                for v in p: counts[v]+=1

#Plotting the color coded nodes
fig, ax = plt.subplots()
unaltered_shortest_paths = counts.values() #List
nodes = G.nodes()
n_color = numpy.asarray([unaltered_shortest_paths[n] for n in range(len(nodes))])
sc = nx.draw_networkx_nodes(G, pos=pos2, node_color=n_color, cmap='gist_heat',
                            with_labels=False, ax=ax, node_size=45)
min_val=int(min(unaltered_shortest_paths))
max_val=int(max(unaltered_shortest_paths))
sc.set_norm(mcolors.Normalize(vmin=0,vmax=max_val))
cbar=fig.colorbar(sc)
cbar.set_label('Number of Shortest Paths')
plt.xlim(-2,N+1,5)
plt.xticks(numpy.arange(0, N, 1))
plt.ylim(N+1,-2,5) 
plt.yticks(numpy.arange(0, N, 1))
plt.axis('on')
title_string=('Stress Centrality (SC)') 
subtitle_string=('Network size: '+str(N)+'x'+str(N))
plt.suptitle(title_string, y=0.99, fontsize=17)
plt.title(subtitle_string, fontsize=9)
plt.xlabel('X')
plt.ylabel('Y')
plt.show()