The Make functor is generic over a field F and a uniform sampler. Genericity over the underlying field is especially useful for testing, using arbitrary-precision rationals.
module F : Intf.Fieldmodule Sampler : sig ... endtype field = F.tThe type of weights.
type particle = P.tThe type of particles.
type 'a t := 'a Stateful_sampling_monad.tstratified_resampling ~target particles computes an array of resampling counts, corresponding to a fresh population. target is the number of particles to resample and particles stores the input population.
val stratified_resampling_list :
target:int ->
(particle -> int -> 'a -> 'a) ->
particle list ->
'a ->
'a tstratified_resampling_list ~target cons pop acc does as stratified_resampling but proceeds on a list of particles instead of an array. The given cons and function is given for each particle the number of copies to be done (which can be 0).
systematic_resampling ~target particles computes an array of resampling counts, corresponding to a fresh population. target is the number of particles to resample and particles stores the input population.
val systematic_resampling_list :
target:int ->
(particle -> int -> 'a -> 'a) ->
particle list ->
'a ->
'a tsystematic_resampling_list ~target cons pop acc does as systematic_resampling but proceeds on a list of particles instead of an array. The given cons and function is given for each particle the number of copies to be done (which can be 0).