Implements drc nonlinear functions into the nlme framework for mixed effects dose-response modelling.
medrm(form, curveid = NULL, data, fct, random, correlation = NULL, weights = NULL, control = NULL, start = NULL, REML = FALSE)
form | Formula describing the dose-response relationship |
---|---|
curveid | Formula with parameter names on the left hand side (divided by +) and a column name in data, denoting a factor, to estimate separate parameters per factor-level. If NULL only fixed effects for a single curve will be estimated. |
data | a data.frame object |
fct | a function of the drc package |
random | a list or one-sided formula describing the random effects |
correlation | additional corClasses object |
weights | additional varClasses object |
control | list with nlme control arguments |
start | optional list with initial values for the fixed components. If NULL the initial values will be found automatically. |
REML | logical value. If TRUE the model is fit by maximizing the restricted log-likelihood, if FALSE log-likelihood is maximized. |
An object of class medrc
An application of medrm is shown on the help pages of data broccoli
. EDx and selectivity indices can be calculated with functions ED
and EDcomp
. Model-averaged ED can be computed by function mmaED
.
Marginal EDx and corresponding function are implemented in functions EDmarg
, using numerical integration conditional on the etimated variance components.