predict.tvcure: Prediction method for tvcure objects

In gwilliford/tvcure: Estimate Proportional Hazards Cure Models with Time-Varying Data

Description

Computes and plots predicted survival curves and cure probabilities at specified values of independent variables

Usage

## S3 method for class 'tvcure'
predict(
  model,
  type = c("basesurv", "suncure", "spop", "uncureprob"),
  variable = NULL,
  values = NULL,
  newX = NULL,
  newZ = NULL,
  xlab = "Time",
  ylab = NULL,
  legendtitle = NULL,
  internals = F
)

Arguments

model	A model of class tvcure
type	The desired quantity to compute. "basesurv" returns the baseline survivor function based on the observed values of the parameters. "suncure" returns the conditional baseline survival function, while "spop" returns the population (i.e., unconditional) survival function. "Uncureprob" returns the probability that an observation is susceptible to failure (i.e., is not cured.)
variable	A string containing the name of the variable to plot predicted values for
values	A vector of values of variable at which survival curves should be plotted
newX	A matrix containing values of hazard variables to calculate predictions for
newZ	A matrix containing values of cure variables to calcluate predictions for
xlab	A label for the x-axis
ylab	A label for the y-axis
legendtitle	Title for the plot legend
internals	A logical value indicating whether the predictions should be returned. If FALSE, only the graph will be returned.
Logical	value indicating whether graphs should be printed in black and white.

Details

Values to compute survival curves and cure probabilities for may be specified in one of two ways. First, specifying the variable and values arguments will compute a separate prediction for all values of the specified variable, with all other variables set to their median. Second, for more complex predictions, matrices containing new values of the hazard and cure variables may be passed using newX and newZ. Values contained in newX and newZ will override any specified variables and values if both are included. By default, the predict function returns a plot of the desired quantity. To access the underlying quantities estimated, set internals = TRUE.

gwilliford/tvcure documentation built on Dec. 20, 2021, 1:51 p.m.