riskCalibrationPlot.default: Calibration plot

In ProfessionalFarmer/loonR: Common function in bioinformatics analysis

Calibration plot

Description

Calibration plot

Usage

## Default S3 method:
riskCalibrationPlot(
  group,
  pred,
  rms.method = FALSE,
  title = "Calibration plot",
  show.oberved.ci = FALSE,
  bins = 10,
  color = "npg",
  ticks.unit = 0.25,
  full.range = TRUE,
  smooth.method = "loess"
)

Arguments

group	Must be a TRUE/FALSE factor
pred	predicted probability
rms.method	If TRUE, use rms::val.prob function instead
bins	Number of bins. Default 20
color	Default npg
ticks.unit	0.25 seq(0, 1, by = 0.25)
full.range	Default TRUE. loess smoothing between 0-1 or first bin to last bin
smooth.method	Smoothing method (function) to use, accepts either NULL or a character vector, e.g. "lm", "glm", "gam", "loess" or a function, e.g. MASS::rlm or mgcv::gam, stats::lm, or stats::loess. "auto" is also accepted for backwards compatibility.

Examples

data(BreastCancer)
BreastCancer = BreastCancer[,-c(1)]
BreastCancer = na.omit(BreastCancer)
m <- glm(Class ~ ., data = BreastCancer, family = binomial)
BreastCancer$pred <- predict(m, type = "response")
riskCalibrationPlot(factor(BreastCancer$Class=="malignant", levels=c(FALSE, TRUE)),
                   BreastCancer$pred)

data(LIRI)

d1 <- LIRI[,-c(1,5)]
m <- glm(status ~ ., data = d1, family = binomial(logit))
d1$pred <- predict(m, type = "response")
loonR::riskCalibrationPlot(factor(LIRI$status), d1$pred)

ProfessionalFarmer/loonR documentation built on Oct. 9, 2024, 9:56 p.m.