<scp>Bilby</scp>-MCMC: an MCMC sampler for gravitational-wave inference

We introduce Bilby-MCMC, a Markov-Chain Monte-Carlo sampling algorithm tuned for the analysis of gravitational waves from merging compact objects. Bilby-MCMC provides parallel-tempered ensemble Metropolis-Hastings sampler with access to block-updating proposal library including problem-specific and machine learning proposals. demonstrate that proposals can produce over 10-fold improvement in efficiency by reducing autocorrelation time. Using variety standard tests, we validate ability independent posterior samples estimate Bayesian evidence. Compared widely-used dynesty nested algorithm, is less efficient producing accurate its estimation However, find drawn are more robust: never failing pass our validation tests. Meanwhile, fails difficult-to-sample Rosenbrock likelihood test, constraining posterior. For CBC problems, this highlights importance cross-sampler comparisons ensure results robust error. Finally, be embarrassingly asynchronously parallelised making it highly suitable wall-time using High Throughput Computing environment. may useful tool rapid gravitational-wave signals during advanced detector era expect have utility throughout astrophysics.