I've computed a distance matrix and I'm trying two approach to visualized it. This is my distance matrix:
delta =
[[ 0. 0.71370845 0.80903791 0.82955157 0.56964983 0. 0. ]
[ 0.71370845 0. 0.99583115 1. 0.79563006 0.71370845
0.71370845]
[ 0.80903791 0.99583115 0. 0.90029133 0.81180111 0.80903791
0.80903791]
[ 0.82955157 1. 0.90029133 0. 0.97468433 0.82955157
0.82955157]
[ 0.56964983 0.79563006 0.81180111 0.97468433 0. 0.56964983
0.56964983]
[ 0. 0.71370845 0.80903791 0.82955157 0.56964983 0. 0. ]
[ 0. 0.71370845 0.80903791 0.82955157 0.56964983 0. 0. ]]
Considering labels from 1 to 7, 1 is really close to 6 and 7 and farther form 4.
At first I tried to use the tSNE dimensionality reduction:
from sklearn.preprocessing import normalize
from sklearn import manifold
from matplotlib import pyplot as plt
from matplotlib.lines import Line2D
import numpy
model = manifold.TSNE(n_components=2, random_state=0, metric='precomputed')
coords = model.fit_transform(delta)
cmap = plt.get_cmap('Set1')
colors = [cmap(i) for i in numpy.linspace(0, 1, simulations)]
plt.figure(figsize=(7, 7))
plt.scatter(coords[:, 0], coords[:, 1], marker='o', c=colors, s=50, edgecolor='None')
markers = []
labels = [str(n+1) for n in range(simulations)]
for i in range(simulations):
markers.append(Line2D([0], [0], linestyle='None', marker="o", markersize=10, markeredgecolor="none", markerfacecolor=colors[i]))
lgd = plt.legend(markers, labels, numpoints=1, bbox_to_anchor=(1.17, 0.5))
plt.tight_layout()
plt.axis('equal')
plt.show()
This produces this plot:
Where we can see that this doesn't show 1 is close to 6 and 7. Instead, it is closest to 4.
Then, not sure if the reduction had stopped at some local minima, I tried to draw a graph:
import networkx as nx
plt.figure(figsize=(7, 7))
dt = [('len', float)]
A = delta
A = A.view(dt)
G = nx.from_numpy_matrix(A)
pos = nx.spring_layout(G)
nx.draw_networkx_nodes(G, pos, node_color=colors, node_size=50)
lgd = plt.legend(markers, labels, numpoints=1, bbox_to_anchor=(1.17, 0.5))
plt.tight_layout()
plt.axis('equal')
plt.show()
As can be seen, the same occurs. If I keep repeating this latest method, I can end up with different sort of graphs:
Here, I get more closer of what I would expect. However, any of these behaviors seems to be right. The distance should be respected no matter how different is the initialization of the graph.
So, I'm wondering what I'm missing in to achieve a good representation of this distance matrix.
Thanks.
