Bed Nets

library(ramp.xds)
library(ramp.control)

Every bed net model has six configurable functions, each one dispatched and configured by a different object. These get called at two different points:

• VectorControl dispatches on class(model$vector_control) == "dynamic"

  • BedNet dispatches on class(model$bednets) == "dynamic"

    • DistributeBedNets dispatches on class(model$bednets$distribute) is there to model mass distribution of ITNs that should sharply increase bednet ownership

    • OwnBedNets dispatches on class(model$bednets$own) is a model that can account for background net replacement and loss

    • UseBedNets dispatches on class(model$bednets$use) is a model for bed net usage

    • BedNetEffects dispatches on class(model$bednets$effects) is a function that modifies search weights

• VectorControlEffectSizes dispatches on class(model$bednets) == "dynamic"

  • BedNetCoverage dispatches on class(model$bednets$coverage[[s]]) for the $s^{th}$ vector species

  • BedNetEffectSizes dispatches on class(model$bednets$effect_size[[s]]) for the $s^{th}$ species

To set all this up, we call a function with a name that calls a setup function and as set of options to set it up, passed as a list:

setup_bednet = function(pars, 
    dist_name = `none`, distribute = list(),
    own_name = `none`, own = list(), 
    use_name = `none`, use = list(), 
    effects_name = `none`, effects = list(), 
    coverage_name = `none`, coverage = list(),
    es_name = `none`, effect_sizes =list()
)

In the simplest models, we can simply set coverage

mod <- xds_setup(MYZname = "SI")

cov_opts <- list(
  F_season = function(t)
    {ifelse(t < 0, 0, (sin(2*pi*(t-365/4) / 365) + 1))}
)

tt = seq(0:730)
with(cov_opts, plot(tt, F_season(tt), type = "l"))

mod <- setup_bednets(mod,
     coverage_name = "func", coverage_opts = cov_opts, 
     effectsizes_name = "lemenach")

tt = seq(0:730)
with(mod$bednets$coverage, plot(tt, F_season(tt), type = "l"))