Simulates dose-response datasets using parametric or non-parametric methods and estimates effective doses (ED values) from each simulated dataset. Useful for assessing the performance of ED estimation methods via Monte Carlo simulation.
Usage
simFct(
noSim,
edVal = c(10, 20, 50),
type = c("non-parametric", "parametric"),
response = c("bin", "con"),
fct = LL.2(),
coefVec,
method = c("sp", "p", "np"),
doseVec,
nVec,
pVec,
rVec,
resVar,
pfct = fct,
reference = NULL,
span = NA,
minmax = "response",
lower = NULL,
upper = NULL,
seedVal = 200810201
)Arguments
- noSim
integer. Number of simulations to run.
- edVal
numeric vector of ED levels to estimate (default is
c(10, 20, 50)).- type
character string. Either "non-parametric" or "parametric" simulation.
- response
character string. Either "bin" (binomial) or "con" (continuous) response.
- fct
dose-response function used for simulation (default is
LL.2()).- coefVec
numeric vector of model coefficients for parametric simulation.
- method
character string. Estimation method: "sp" (semi-parametric), "p" (parametric), or "np" (non-parametric).
- doseVec
numeric vector of dose values.
- nVec
numeric vector of sample sizes per dose (for binomial response).
- pVec
numeric vector of expected response probabilities (for non-parametric simulation).
- rVec
numeric vector of responses.
- resVar
numeric. Residual variance (for continuous response).
- pfct
dose-response function used for fitting (defaults to
fct).- reference
character string specifying the reference for ED estimation.
- span
numeric. Smoothing parameter for local regression. NA uses default.
- minmax
character string. Type of min/max calculation. Default is "response".
- lower
numeric. Lower bounds for optimization.
- upper
numeric. Upper bounds for optimization.
- seedVal
integer. Random seed for reproducibility (default is 200810201).