search

scramblingbalam / IIT_in_orgs

Applying Information Integration Theory
0 stars 0 forks source link

Try creating a simulated network in iGraph #5

Closed scramblingbalam closed 6 years ago

scramblingbalam commented 7 years ago

Try creating a simulated network with the attributes:

100 nodes 1000 edges Reciprocity=0.5 Clustering=0.3 Edges with random weights, I think range(0,1)

scramblingbalam commented 7 years ago

In the create_random function, I forgot that random.choice(all_possible_edges) is with replacements so I was getting repeat edges which I guess iGraph supports. To solve this I will change to numpy.random.choice(all_possible_edges,replace =False) https://docs.scipy.org/doc/numpy/reference/generated/numpy.random.choice.html

scramblingbalam commented 7 years ago

Numpy doesn't work since it see the list of edges as a 2D array so going to go back to the random package and use random.sample(all_possible_edges,m)

scramblingbalam commented 7 years ago

So with the basic create_random function working and a realization that the find_all_paths function won't work at scale I'm re-scoping to a network one order of magnitude smaller because I can no longer have number of edges and number of nodes as a metric I will replace them with density. so the new list of attributes is: Nodes = 20-30 Note basically the biggest network it will run at. Density 0.10101010101010101 Reciprocity=0.5 Clustering=0.3 Edges with random weights, I think range(0,1)