nCal (version 0.3-2)

drm.fit: Fit drm

Description

drm.fit fit concentration-response curves using drm function from drc package.

Usage

drm.fit (formula, data, robust="mean", fit.4pl=FALSE, force.fit=TRUE, 
    weighting=FALSE, coln.weight=NULL, pow.weight=1, verbose=FALSE) 

getVarComponent.drc (fit)

Arguments

formula

a formula object.

data

a data frame object.

robust

a string. Passed to drm. See ?drm for more details.

fit.4pl

boolean. If TRUE, 4PL model is fitted. If FALSE, 5PL model is fitted.

force.fit

boolean. If TRUE, a fit is returned even if it is determined that the goodness of fit is poor.

weighting

boolean. If TRUE, weighted least square optimization is carried out. The weights are in the form of coln.weight^pow.weight.

coln.weight

string. The column used in the weights. If NULL, default to the response variable in the formula.

pow.weight

numeric. The power used in the weights.

verbose

Boolean. If TRUE, verbose messages are printed.

fit

a drm object.

Value

• An object of type drm.

Details

drm.fit differs from drc::drm in several aspects. (1) It tries several self start functions in order to get better fits. (2) It uses gof.threshold to report lack of fit. (3) It tried to determine whether the standard deviation of the parameter estimates can be estimated. (4) It implements weighting through coln.weight and pow.weight.

Examples

# simulate a dataset
set.seed(1)
log.conc=log(1e4)-log(3)*9:0
n.replicate=2
fi=simulate1curve (p.eotaxin[1,], rep(log.conc,each=n.replicate), sd.e=0.2)
dat.std=data.frame(fi, expected_conc=exp(rep(log.conc,each=n.replicate)), analyte="Test", 
    assay_id="Run 1", sample_id=NA, well_role="Standard", dilution=rep(3**(9:0), each=n.replicate),
    replicate=rep(1:n.replicate, 10))

fit = drm.fit(fi ~ expected_conc, data = dat.std, force.fit=TRUE)
plot(fit, log="xy")
fit

fit.2 = drm.fit(fi ~ expected_conc, data = dat.std, force.fit=TRUE, weighting=TRUE, pow.weight=-1)
plot(fit.2, log="xy")
fit.2