Learn R Programming

drc (version 0.8-2)

bindrc: Fitting multiple dose-response curves for binomial data

Description

'bindrc' fits logit/probit regression models to multiple dose-response curves. The models are linear in the logarithm of dose in the logit/probit scale.

Usage

bindrc(formula, weights, curve, int, slope, data = NULL, na.action = na.fail, 
link = "logit", log =TRUE, startVal, lower, upper, fixedLow = NULL, fixedUp = NULL, 
names = c("a", "b", "c", "d"))

Arguments

formula
a symbolic description of the model to be fit in the form 'response $~$ dose'.
weights
a numeric containing number of trials per experiment, eg per dose level.
curve
a numeric vector or factor containing the grouping of the data into separate curves. If not specified all observations are assumed to belong to a single curve!
int
a numeric vector or factor containing the grouping of the data for the intercept.
slope
a numeric vector or factor containing the grouping of the data for the intercept.
data
an optional data frame containing the variables in the model.
na.action
a function which indicates what should happen when the data contain 'NA's. The default is 'na.fail'. To omit 'NA's use 'na.omit'.
link
a character string specifying the model: "logit", "probit" or "cloglog".
log
logical. If TRUE (default) independent variable is log-transformed. If FALSE no log-transformation is used.
startVal
an optional numeric vector containing start values for all parameters in the model. Overrules any self starter facility.
lower
a numeric vector or factor containing the grouping of the data for the lower limit.
upper
a numeric vector or factor containing the grouping of the data for the upper limit.
fixedLow
a vector for fixing the lower limit for some curves.
fixedUp
a vector for fixing the upper limit for some curves.
names
a character vector with the names of the parameters.

Value

  • An object of class 'drc'. Various extractors exist for objects of this class.

Details

The model fit has the form $$p = c + (d-c) F( \alpha + \beta \log dose)$$ for each curve with F the inverse of logit or probit. Maximum likelihood estimation is employed.

See Also

See multdrc for fitting data with continuous response.