I've set up the following code to read in a .graphml file, preform a calculation (eigenvalues) and then plot the results. Here is the code I have so far:
import numpy as np
import networkx as nx
import matplotlib.pyplot as plt
# Read in the Data
G = nx.read_graphml("/home/user/DropBox_External_Datasets/JHU_Human_Brain/cat_brain_1.graphml")
nx.draw(G)
plt.savefig("test_graph.png")
Z = nx.to_numpy_matrix(G)
# Get Eigenvalues and Eigenvectors
# ----------------------------------------------------------------------------------
#
e_vals, e_vec = np.linalg.eigh(Z)
print("The eigenvalues of A are:", e_vals)
print("The size of the eigenvalues matrix is:", e_vals.shape)
# ----------------------------------------------------------------------------------
plt.plot(e_vals, 'g^')
plt.ylabel('Eigenvalues')
# plt.axis([-30, 300, -15, 30]) # Optimal settings for Rhesus data
# plt.axis([-0.07, 1, -0.2, 1.2]) # range to zoom in on cluster of points in Rhesus data
plt.grid(b=True, which='major', color='b', linestyle='-')
plt.show()
But no gridlines or axes show up on the graph. Is there something other then plt.grid() that I need to use?
I have been finding that using the matplotlib object oriented API is a more robust way to make things work as expected. Pyplot is essentially a big wrapper for the object-oriented calls. I've written something that should be equivalent:
import matplotlib.pyplot as plt
# ... your other code here
# Using subplots
fig, ax = plt.subplots(ncols=1, nrows=1) # These arguments can be omitted for one
# plot, I just include them for clarity
ax.plot(e_vals, 'g^')
ax.set_ylabel('Eigenvalues')
ax.grid(b=True, which='major', color='b', linestyle='-')
plt.show()