Experiments benchmarking the task of loading a graph, executing a first-order random walk and a second-order random walk on a number of graph from the literarure using Ensmallen and other libraries from the literature.
To reproduce the benchmarks just clone the repository and run the run_tests.sh bash script.
Do note that the results reported are of an Ensmallen version of about 1 year ago, and the more recent versions are considerably faster. We are still striving to achieve better throuput for these tasks, but our focus has currently moved to the validation of real-world datasets and identification of significant biases and oddities that may be present therein.
