Come analyze HEASARC, IRSA, and MAST data in the cloud! The Fornax Initiative is now welcoming all interested beta users.
Goodman–Weare.
Same Hessian-based proposal; the walker ensemble (10 by default, raised to 32 in the example below) takes care of step scaling via the stretch move, so the proposal width matters less here than for MH.
chain proposal gaussian fit chain type gw chain length 5000 chain burn 1000 chain walkers 32 parallel walkers 4 chain run gw.fits
Observed: 9 s elapsed time (the 32-walker ensemble parallelizes efficiently across 4 cores).