Besides the rates that mizer calculates itself, a model can carry extra
contributions to the mortality rate and to the encounter rate. Each
contribution is an R function, registered here by name, that mizer calls at
every time step: getMort() adds the result of every function listed in
other_mort(params) to the mortality rate and getEncounter() adds the
result of every function listed in other_encounter(params) to the encounter
rate.
Usage
other_mort(params)
other_mort(params) <- value
other_encounter(params)
other_encounter(params) <- valueValue
A named list with the names of the functions contributing to the rate, excluding any that belong to a component.
Details
Use these when the extra contribution depends on the state of the model, as
a starvation mortality or a density-dependent mortality does. A contribution
that is simply a fixed array is better set with ext_mort() or
ext_encounter(), which is what mizer's own external mortality and external
encounter use.
Assigning NULL to an entry removes that contribution:
How your function is called
Each registered function is called as
and has to return an array with the same dimensions as the rate it
contributes to, that is species x size. The component argument holds the
name under which the function is registered, so a single implementation can
serve several entries. Always give your function a ... argument so that it
tolerates being passed arguments it does not use.
Make sure the function depends continuously on the abundances, for the
reasons set out in setRateFunction().
Contributions belonging to a component
A component set up with setComponent() can contribute to these rates
through that function's mort_fun and encounter_fun arguments. Such an
entry belongs to its component: setComponent() sets it, removeComponent()
removes it and getComponent() reports it. The accessors here therefore
leave those entries alone. other_mort() and other_encounter() list only
the contributions that do not belong to a component, and assigning through
them preserves the ones that do. This mirrors other_params(), which
likewise hides the parameters belonging to a component, and it makes it
impossible to wipe out a component's contribution by assigning a whole list.
Examples
# An extra mortality that grows with the total biomass in the community.
# Like any custom rate function it has to live in the global environment
# or in a package, so that mizer can find it by name.
crowdingMort <- function(params, n, n_pp, n_other, t, component, ...) {
biomass <- sum(n %*% (params@w * params@dw))
# Same dimensions as the mortality rate: species x size
0 * params@mu_b + 1e-14 * biomass
}
params <- NS_params
other_mort(params)[["crowding"]] <- "crowdingMort"
#> Error in setRateContributions(params, "other_mort", value, noun = "mortality", arg = "mort_fun"): The entry `crowding` of `other_mort` needs to be the name of a function.
other_mort(params)
#> list()
# The contribution is included in the mortality rate
range(getMort(params) - getMort(NS_params))
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> ℹ No `a` column so using a = 0.01 in w = a l^b, with w in g and l in cm.
#> ℹ No `b` column so using the isometric default b = 3 in w = a l^b.
#> [1] 0 0
# and can be removed again
other_mort(params)[["crowding"]] <- NULL
other_mort(params)
#> list()