probably-baysig
probably-baysig copied to clipboard
discrete parameters
if we modify the type of runChain such that the posterior function has type Vector Double -> Vector Int -> Double (but the gradient has the same type, or maybe it still takes the Vector Int as an argument but of course it cant produce a gradient on discrete parameters so it just treats discrete parameters as fixed).. then... do something..
we could just do #8 first, then modify the Haario adaptive metropolis strategy to work on both Ints and Reals. Then at least we have adaptive metropolis for integer parameters.