Create observer

Create an observer to extract additional details from your model during the sampling process.

Usage

monty_observer(observe, finalise = NULL, combine = NULL, append = NULL)

Arguments

observe

A function that will run with arguments model (the model that you passed in to monty_model) and rng (an rng object). This function should return a list. It is best if the list returned is named, with no duplicated names, and with return values that have the same exact dimensions for every iteration. If you do this, then you will not have to provide any of the following arguments, which are going to be hard to describe and worse to implement.

finalise

A function that runs after a single chain has run, and you use to simplify across all samples drawn from that chain. Takes a single argument which is the list with one set of observations per sample.

combine

A function that runs after all chains have run, and you use to simplify across chains. Takes a single argument, which is the list with one set of observations per chain.

append

A function that runs after a continuation of chain has run (via monty_sample_continue. Takes two arguments representing the fully simplified observations from the first and second chains.

Value

An object with class monty_observer which can be passed in to monty_sample.

Details

Sometimes you want to extract additional information from your model as your chain runs. The case we see this most is when running MCMC with a particle filter (pmcmc); in this case while the likelihood calculation is running we are computing lots of interesting quantities such as the final state of the system (required for onward simulation) and filtered trajectories through time. Because these are stochastic we can't even just rerun the model with our sampled parameter sets, because the final states that are recovered depend also on the random number generators (practically we would not want to, as it is quite expensive to compute these quantities).

The observer mechanism allows you to carry out arbitrary additional calculations with your model at the end of the step.