utils.py

#!/usr/bin/env python
# coding: utf-8

# In[ ]:

import torch
import networkx as nx
import numpy as np
import pandas as pd


# In[ ]:


def create_Graphs_with_attributes(edgelist_filepath,attributes_filepath):
    graph=nx.read_edgelist(edgelist_filepath,nodetype=int)
    attributes=pd.read_csv(attributes_filepath,index_col=['node'])
    att_values = {a:{'role':b[0],'community':b[1]} for a,b in enumerate(attributes.values)}
    nx.set_node_attributes(graph,att_values)
    return graph


# In[ ]:


def create_train_test(graph):
    X_train,Y_train,X_test,Y_test=[],[],[],[],[]
    for node, data in graph.nodes(data=True):
        if data['role'] in ['Administrator','Instructor']:
            X_train.append(node)
            Y_train.append(data['role']=='Administrator')
        elif data['role'] =='Member':
            X_test.append(node)
            Y_test.append(data['community']=='Administrator')
    return np.asarray(X_train),np.asarray(Y_train),np.asarray(X_test),np.asarray(Y_test)



def create_features(graph): # create input features, concatenation of identity matrix and shortest paths to targets
    A = nx.to_numpy_matrix(graph)
    X_1 = torch.eye(A.shape[0]) # identity matrix
    X_2 = torch.zeros((A.shape[0], 2)) # shortest path to the targets as 2nd input feature
    node_distance_instructor = nx.shortest_path_length(graph, target=33)
    node_distance_administrator = nx.shortest_path_length(graph, target=0)

    for node in graph.nodes():
        X_2[node][0] = node_distance_administrator[node]
        X_2[node][1] = node_distance_instructor[node]

    return torch.cat((X_1,X_2),dim=1)