getTauSurface: Compute interaction index surface and confidence intervals

In drugCombo: Drug Interaction Modeling Based on Loewe Additivity Following Harbron's Approach

Description

Computes estimates and confidence intervals for the interaction surface for all dose combinations.

Usage

getTauSurface(
  fit,
  data = NULL,
  addCI = TRUE,
  method = c("default", "boot"),
  level = 0.95,
  niter = 100,
  resample = c("all", "mono", "stratified"),
  seed = NULL,
  ...
)

Arguments

fit	A HarbronFit object returned by fitModel.
data	Dose combinations to compute interaction index for. If NULL (default), taken from the fit object.
addCI	Whether confidence intervals need to be computed.
method	Which method to use to calculate confidence intervals: "default" for Wald-type or "boot" for non-parametric bootstrap.
level	The confidence level required for the confidence intervals (default is 0.95).
niter	Number of bootstrap samples to use.
resample	Resampling method for bootstrap. Either "all" (default) for resampling from all data, "mono" for separately resampling monotherapy and combination data, or "stratified" for resampling at each dose combination separately. Note that the latter method is not meaningful if there are no replicates in the data.
seed	Random seed to use for bootstrap
...	Further parameters that are passed to getBootTaus.

Value

An object of class "tauSurface" which is essentially a list with the following components: data frame with interaction index (tau) estimates, standard errors and point-wise confidence intervals, formulas for computing tau at any given dose (only for models with continuous functions used to define tau), and details on the tau specification from the fit. In addition, if the "boot" method was used, all the bootstrap estimates are returned and can be accessed with bootstrapCoefs.

Author(s)

Maxim Nazarov

See Also

plot.tauSurface, contour.tauSurface for visual representation of the tau surface.

drugCombo documentation built on June 30, 2021, 1:07 a.m.