cut_perf {BioPred}R Documentation

Cutoff Performance Evaluation

Description

This function evaluates the performance of a predictive model at a selected cutoff point.

Usage

cut_perf(
  yvar,
  censorvar = NULL,
  xvar,
  cutoff,
  dir,
  xvars.adj = NULL,
  data,
  type,
  yvar.display = yvar,
  xvar.display = xvar
)

Arguments

yvar

Response variable name.

censorvar

Censoring variable name (0-censored, 1-event).

xvar

Biomarker name.

cutoff

Selected cutoff value.

dir

Direction for desired subgroup (">", ">=", "<", "<=").

xvars.adj

Other covariates to adjust when evaluating the performance.

data

Data frame containing the variables.

type

Type of analysis: "c" for continuous, "s" for survival, and "b" for binary.

yvar.display

Display name of response variable.

xvar.display

Display name of biomarker variable.

Value

A list containing various performance metrics and optionally, plots.

Examples

# Load a sample dataset
data <- data.frame(
  survival_time = rexp(100, rate = 0.1),  # survival time
  status = sample(c(0, 1), 100, replace = TRUE),  # censoring status
  biomarker = rnorm(100, mean = 0, sd = 1),  # biomarker levels
  covariate1 = rnorm(100, mean = 50, sd = 10)  # an additional covariate
)
# Perform cutoff performance evaluation for continuous outcome
data$continuous_outcome <- rnorm(100, mean = 10, sd = 5)
cut_perf(
  yvar = "continuous_outcome",
  xvar = "biomarker",
  cutoff = 0.5,
  dir = ">=",
  data = data,
  type = "c",
  yvar.display = "Continuous Outcome",
  xvar.display = "Biomarker Level"
)

# Perform cutoff performance evaluation for binary outcome
data$binary_outcome <- sample(c(0, 1), 100, replace = TRUE)
cut_perf(
  yvar = "binary_outcome",
  xvar = "biomarker",
  cutoff = 0,
  dir = "<=",
  data = data,
  type = "b",
  yvar.display = "Binary Outcome",
  xvar.display = "Biomarker Level"
)

[Package BioPred version 1.0.2 Index]