Bayesian inference on network properties with partial crawl of the network using random walks.
Software of the research paper:
"Inference in OSNs via Lightweight Partial Crawls" by Konstantin Avrachenkov, Bruno Ribeiro and Jithin K. Sreedharan
to appear in the proceedings of ACM SIGMETRICS/PERFORMANCE 2016.

"friendster_community1_trimmed.edgelist" is a largest connected component extracted from one subgraph of Friendster. Data is collected from https://snap.stanford.edu/

The file "HypRW_friendster.py" is the script which does the random walk crawling and plots the approximate posterior distribution and the empirical distribution.

The Python script requires various modules like NetworkX, Scipy and NumPy.