Skip to contents

Likelihood-ratio tests for dose-related parameters

dose_lrt.Rd

Compute global and individual likelihood-ratio tests for dose-related parameters in supported amerasfit methods.

Usage

dose_lrt(object, ...)

# S3 method for class 'amerasfit'
dose_lrt(
  object,
  methods = c("RC", "ERC", "MCML"),
  type = c("global", "individual"),
  data = NULL,
  ...
)

Arguments

object

An object of class amerasfit.

methods

Character vector of fitted methods to test. Supported methods are "RC", "ERC", and "MCML".

type

Character vector selecting which tests to return. "global" tests all dose-related parameters jointly. "individual" tests each dose-related parameter separately. By default, both are returned.

data

Optional data frame required when object was fitted with keep.data=FALSE.

...

Currently unused.

Value

A data frame with one row per likelihood-ratio test and columns:

method

The ameras estimation method.

type

Either "global" or "individual".

term

The tested dose term. For global tests this is "all dose terms".

df

Degrees of freedom for the likelihood-ratio test.

logLik

The fitted model log likelihood.

logLik.null

The constrained null-model log likelihood.

statistic

The likelihood-ratio statistic.

p.value

The chi-square reference p-value.

null.optim.convergence

Optimizer convergence code for the constrained null fit. A warning is issued if this code is nonzero.

Details

dose_lrt() computes likelihood-ratio tests for dose-related parameters in fitted RC, ERC, and MCML models. The null hypothesis sets the tested dose-related parameter(s) equal to zero. FMA and BMA are not supported by this function.

For transformed fits, nuisance parameters are evaluated through the fitted transformation. The tested dose parameters are then fixed to zero before the likelihood is evaluated. This is most straightforward for component-wise transformations, including the default transformations used by ameras. Results from coupled custom transformations should be interpreted with caution. For example, if a custom transformation defines one reported parameter as a function of multiple optimizer-scale parameters, or makes a non-dose reported parameter depend directly on a dose parameter, fixing only the reported dose parameter to zero may not correspond to the intended constrained model.

For speed, constrained null fits report only the optimizer convergence code. They do not run the additional numerical-gradient diagnostics stored for the original RC, ERC, and MCML fits.

Examples

set.seed(1)
dat <- data.frame(
  Y = 1 + 0.8 * seq(-1, 1, length.out = 30) + rnorm(30, sd = 0.2),
  D = seq(-1, 1, length.out = 30)
)

fit <- ameras(Y ~ dose(D), data = dat, family = "gaussian",
              methods = "RC", print = FALSE)
#> Fitting RC

dose_lrt(fit)
#>   method       type           term df   logLik logLik.null statistic
#> 1     RC     global all dose terms  1 8.776784   -21.27384  60.10124
#> 2     RC individual           dose  1 8.776784   -21.27384  60.10124
#>        p.value null.optim.convergence
#> 1 9.010161e-15                      0
#> 2 9.010161e-15                      0
dose_lrt(fit, type = "global")
#>   method   type           term df   logLik logLik.null statistic      p.value
#> 1     RC global all dose terms  1 8.776784   -21.27384  60.10124 9.010161e-15
#>   null.optim.convergence
#> 1                      0