| initialize-methods {simecol} | R Documentation |
This function is used to initialize objects derived from the
simObj superclass, it is by default automatically called during
object creation and by sim.
## S4 method for signature 'simObj': initialize(.Object, ...)
.Object |
simObj instance which is to be
re-initialized.
|
... |
provided for compatibility with the default method of
initialize, or slots of the object which is to be created (in
case of new).
|
new.initialize function is normally called implicitly by
new to create new objects. It may also be called explicitly
to return a cloned and re-initialized object.
The simecol version of initialize provides an
additonal mechanism to call a user specified function provided in
the initfun slot of a simObj instance that can
perform computations during the object creation process. The
initfunc must have obj as its only argument and must
return the modified version of this obj, see examples
below. As a side effect end to ensure consistency,
initialize clears outputs stored in slot out from
former simulations.
## Note: new calls initialize and initialize calls initfunc(obj)
lv_efr <- new("odeModel",
main = function (time, init, parms, ...) {
x <- init
p <- parms
S <- approxTime1(inputs, time, rule=2)["s.in"]
dx1 <- S * p["k1"] * x[1] - p["k2"] * x[1] * x[2]
dx2 <- - p["k3"] * x[2] + p["k2"] * x[1] * x[2]
list(c(dx1, dx2))
},
parms = c(k1=0.2, k2=0.2, k3=0.2),
times = c(from=0, to=100, by=0.5),
init = c(prey=0.5, predator=1),
solver = "rk4",
initfunc = function(obj) {
tt <- fromtoby(times(obj))
inputs(obj) <- as.matrix(data.frame(
time = tt,
s.in = pmax(rnorm(tt, mean=1, sd=0.5), 0)
))
obj
}
)
plot(sim(lv_efr)) # initialize called automatically
plot(sim(lv_efr)) # automatic initialization, different figure
lv_efr<- initialize(lv_efr) # re-initialize manually
plot(sim(lv_efr, initialize = FALSE)) # simulation with that configuration