Visualizing networks with homophily and two groups of different sizes

The visualisation of a social network which is based on Karimi et al. 2017 model. In the model, it is assumed that there is an underlying homophily parameter due to node attributes which leads to the formation of links between members of each group in the network. The model considers the preferential attachment property proposed by Barabási and Albert. For the implementation of the model in python, you can visit this repository.

For users who are not familiar by how to render html files, here is the solution:

• Download a copy of an existing Git repository on your machine using git clone https://github.com/neuronphysics/Homophilic-Ntw-Viz.git command.

• Download and install brackets which is an editor for web developers.

• Go to the network project and open the network.html using File> Open Folders in your brackets editor.

• select File > Live Preview or click the lightning bolt icon on the rightside toolbar. Brackets will launch Chrome and open your file in a new tab.

The output should be similar to the following image:

In this model, nodes have only two attributes, so they are assigned to two groups with different sizes. The homophily parameter alters between 0 to 1. Here it was posited that group a is the minority group in the network and group b has the majority. The nodes in the minority group are shown in red.

The parameters that can be changed are:

• Number of nodes - The number of nodes which create the network

• Number of edges of a new node - The minimum number of edges which a node can have in this network

• Minority fraction - The fraction of nodes that belong to the group a

• $h_{ab}$ - The probability of connection between group a and b

• $h_{ba}$ - The probability of connection between the majority group members with the minorities

• Press button generate - Make a new realization of the network with the modified input parameters

The colour bars (the top left side) illustrate the fraction of minority nodes (red) in the top 10% with the highest degree inside the generated network.

Author of the code

• Zahra Sheikhbahaee - PhD in Astrophysics