Code to accompany UAI 2018 paper 'Sampling and Inference for Beta Neutral-to-the-Left Models of Sparse Networks' (supplement).
The Stanford Large Network Dataset Collection contains a number of interesting temporal networks. We recommend preprocessing the datasets as follows:
wget https://snap.stanford.edu/data/$NAME.txt.gz
gunzip $NAME.txt.gz
sort -k3 -n $NAME.txt > sorted-$NAME.txt
See examples/mle.jl for an example of computing MLEs on massive datasets.
See examples/gibbs.jl for an example of performing posterior inference over parameters
and latent variables on datasets of modest size (e.g., hundreds or thousands of nodes).
The code can be run interactively (i.e., section by section) or as a script
from the Julia REPL or command line.
See examples/gibbs_plots.jl for some example plots for assessing sampler output.
See examples/gibbs_ess_experiments.jl for code used to produce the tables in Section 5.1
of the paper.