adaptive HMC

[tpapp]

I wonder what you think about adaptive HMC, in particular Wang, Mohamed, and Freitas (2013). Note that I am not asking you to implement it, just considering adding it myself, however, I am a mere user of MCMC so I would value your insight before I spend time on this.

[papamarkou]

It's an interesting paper @tpapp; apart from HMC and NUTS, a few more very promising Hamiltonian Monte Carlo will be added, currently being developed by Chris Sherlock from Lancaster university, it might be a year before these methods are fully developed though (we are potential project partners with Chris, subject to grant success). Will keep you posted.

[ElleryL]

wondering this is already implemented in Julia