# confint.onls: Confidence intervals for 'onls' model parameters In onls: Orthogonal Nonlinear Least-Squares Regression

 confint.onls R Documentation

## Confidence intervals for 'onls' model parameters

### Description

Computes confidence intervals for one or more parameters of an onls model. As in MASS:::confint.nls, these are based on profile likelihoods, using onls:::profile.onls and onls:::confint.profile.onls.

### Usage

## S3 method for class 'onls'
confint(object, parm, level = 0.95, ...)


### Arguments

 object an object returned from onls. parm a specification of which parameters are to be given confidence intervals, either a vector of numbers or a vector of names. If missing, all parameters are considered. level the confidence level required. ... additional argument(s) for methods.

### Details

Profiling the likelihood uses the following strategy:
If θ is the parameter to be profiled and δ the vector of remaining parameters,
1) compute the log-likelihood of the model \mathcal{L}(θ^{*}, δ^{*}) using the converged parameters,
2) compute a lower bound θ^{*} - 0.6 \cdot σ(θ^{*}) for the lower confidence limit,
3) define a grid of values ranging from θ^{'} to θ^{*} (e.g., 100 equidistant points),
4) for each grid value θ_i, compute the profile log-likelihood value \mathcal{L}_1(θ_i) by maximizing \mathcal{L}(θ_i, δ) over δ-values by fixing θ at θ_i,
5) find the confidence level by interpolation of the profile traces obtained from 4).

### Value

A matrix (or vector) with columns giving lower and upper confidence limits for each parameter. These will be labelled as (1 - level)/2 and 1 - (1 - level)/2 in % (by default 2.5% and 97.5%).

### Author(s)

Andrej-Nikolai Spiess, taken and modified from the nls functions.

### Examples


DNase1 <- subset(DNase, Run == 1)
DNase1$density <- sapply(DNase1$density, function(x) rnorm(1, x, 0.1 * x))
mod1 <- onls(density ~ Asym/(1 + exp((xmid - log(conc))/scal)),
data = DNase1, start = list(Asym = 3, xmid = 0, scal = 1))
confint(mod1)



onls documentation built on Oct. 31, 2022, 5:06 p.m.