ThickAss.py

# %%
import numpy as np
import pandas as pd
import igraph as igraph
import matplotlib.pyplot as plt

#data = pd.read_excel (r'C:\Users\rixtb\Documents\Master\Data analysis\Datasets\manufacturing_emails_temporal_network.xlsx')
#data = pd.read_excel (r'C:\Users\rixtb\Documents\Master\Data analysis\Datasets\oefenset.xlsx')
data = pd.read_excel('manufacturing_emails_temporal_network.xlsx')

# %% A
Nnodes = np.max([data['node1'].max(), data['node2'].max()])
G = data.drop(['timestamp'],axis=1)
B = G.drop_duplicates()
g = igraph.Graph()
g.add_vertices(Nnodes)

Nlinks = len(B)

col1 = B['node1']; col2 = B['node2']
col1 = col1.tolist(); col2 = col2.tolist()

for i in range(Nlinks):
    g.add_edges([(col1[i]-1,col2[i]-1)])

ED = g.degree()
print(ED)
p = g.density(loops=False) #ratio between the actual links and the possible edges
print(p)
average_degree = 2* Nlinks/Nnodes #see slides lecture 1
print(average_degree)
var = np.var(ED)
print(var)

plt.hist(ED,bins=Nnodes)

plt.plot()

pd = g.assortativity_degree()
print(pd)
clust_coef = g.transitivity_undirected()
print(clust_coef)

average_path_length=g.average_path_length()
print(average_path_length)
hopcount_max = np.max(g.shortest_paths())
print(hopcount_max)
#A = (g.get_adjacency())
#print(A)
eigenvalueA= g.eigen_adjacency()
print(eigenvalueA)
    
Q = np.array(g.laplacian(weights=None, normalized=False));
eig = np.linalg.eig(Q)[0];
eig.sort()
print(eig[1])
#print(min(i for i in eig if i > 0))  #first non-zero eigenvalue laplacian

# %% B
tmax = 10 #data('timestamp').max()
A = np.eye(Nnodes,dtype=int)
g = igraph.Graph()
g.add_vertices(Nnodes)

for t in range(tmax): #generating the adjacency matrix
    G = data[data.timestamp==t]

    col1 = G['node1'].to_numpy(); col2 = G['node2'].to_numpy()
    
    for i in range(len(col1)):
        g.add_edges([(col1[i]-1,col2[i]-1)])
        
    Mat=g.get_adjacency()
    Mat=np.array([Mat.data])
    A = np.dot(Mat,A) + np.eye(Nnodes)

# %%
# %whos

# %%
col1

# %%