dcaR Documentation

Calculate Decision Curve Data

Description

Calculate Decision Curve Data

Usage

dca(...)

## S3 method for class 'coxph'
dca(
  ...,
  model.names = do::get_names(...),
  test.harm = 0,
  new.data = NULL,
  times = "median"
)

## S3 method for class 'cph'
dca(
  ...,
  model.names = do::get_names(...),
  test.harm = 0,
  new.data = NULL,
  times = "median"
)

## S3 method for class 'glm'
dca(..., model.names = do::get_names(...), test.harm = 0, new.data = NULL)

## S3 method for class 'lrm'
dca(..., model.names = do::get_names(...), test.harm = 0, new.data = NULL)

Arguments

...

one or more results of logistic or cox regression

model.names

names for models

test.harm

test harm, default is 0

new.data

new data for validation

times

times for cox regresion, default is 'median'

Value

a dataframe contains thresholds, TPR: true positive rate, FPR: false positive rate, NB: net benefit, model: model names.

References

Vickers, A. J., & Elkin, E. B. (2006). Decision Curve Analysis: A Novel Method for Evaluating Prediction Models. Medical Decision Making, 26(6), 565–574. https://doi.org/10.1177/0272989X06295361