Write code to reproduce the following figures comparing Erdos-Renyi random networks to Barabasi-Albert networks. Note that...

Transcribed Image Text:

Write code to reproduce the following figures comparing Erdos-Renyi random networks to Barabasi-Albert networks. Note that Networkx has a function for generating Barabasi-Albert (BA) networks called barabasi_albert_graph (more here). (a) Reproduce the following figure by generating BA networks of various sizes (i.e., varying N). Use the parameter choice m = 2. For each choice of N, consider the average average clustering coeffi- cient of 5 different BA networks. Networkx also has built-in functions for calculating clustering coefficients. Are your results well approximated by (ln(N))²/N as in the figure? 10 10 (C) 10 10⁰ F 10 10 105 104 10³ k 10² 10¹ 10 10⁰ (b) Plot the degree distributions for BA networks with N = m = 1, 3, 5, 7. RANDOM NETWORK ~1/N 10° 10 (c) Reproduce the following figure by growing a single BA network up to 105 nodes (or however many nodes your computer can handle). Use m = 2, but explain what will happen if we vary m. While growing the network, note the nodes added at time t = 1, 10, 10², 10³, 104, 105 and track their respective degrees over time. SINGLE NETWORK B = 1/2 10¹ 10 N 104 10² (ln N)² N ta 10 10³ t 104 10 105 and for the parameter choices 10³ 10 You can use the Networkx function for generating BA networks to check your work, but you may need to write your own code for generating BA networks if you want to track the degree of an individual node over time. This tutorial may be helpful. Write code to reproduce the following figures comparing Erdos-Renyi random networks to Barabasi-Albert networks. Note that Networkx has a function for generating Barabasi-Albert (BA) networks called barabasi_albert_graph (more here). (a) Reproduce the following figure by generating BA networks of various sizes (i.e., varying N). Use the parameter choice m = 2. For each choice of N, consider the average average clustering coeffi- cient of 5 different BA networks. Networkx also has built-in functions for calculating clustering coefficients. Are your results well approximated by (ln(N))²/N as in the figure? 10 10 (C) 10 10⁰ F 10 10 105 104 10³ k 10² 10¹ 10 10⁰ (b) Plot the degree distributions for BA networks with N = m = 1, 3, 5, 7. RANDOM NETWORK ~1/N 10° 10 (c) Reproduce the following figure by growing a single BA network up to 105 nodes (or however many nodes your computer can handle). Use m = 2, but explain what will happen if we vary m. While growing the network, note the nodes added at time t = 1, 10, 10², 10³, 104, 105 and track their respective degrees over time. SINGLE NETWORK B = 1/2 10¹ 10 N 104 10² (ln N)² N ta 10 10³ t 104 10 105 and for the parameter choices 10³ 10 You can use the Networkx function for generating BA networks to check your work, but you may need to write your own code for generating BA networks if you want to track the degree of an individual node over time. This tutorial may be helpful.

Answer rating: 100% (QA)

a Here is a code in Python that generates and plots BA and ErdsRnyi random networks and calculates their average clustering coefficients import networkx as nx import matplotlibpyplot as plt import num... View the full answer