Prediction Model Tools

Provides additional functions for evaluating predictive models, including plotting calibration curves and model-based Receiver Operating Characteristic (mROC) based on Sadatsafavi et al (2021) .

Overview

predtools provides miscellaneous tools for developing and evaluating prediction models.

Table of Contents

• Installation
• Example
• Model-based ROC
• Intercept Adjustment
• Calibration Plot
• Unit Normal Loss Integral in Two Dimensions

Installation

You can install the released version of predtools from CRAN with:

install.packages("predtools")

And the development version from GitHub with:

# install.packages("remotes")
remotes::install_github("resplab/predtools")

Example

The function calibration_plot takes observed and predicted values from a prediction model and uses ggplot2 to produce a calibration plot:

library(predtools)
library(dplyr)
x <- rnorm(100, 10, 2)
y <- x + rnorm(100,0, 1)
data <- tibble(x,y)
calibration_plot(data, obs = "x", pred_1 = "y")

See vignettes for more advanced functionalities, including model-based ROC, intercept adjustment, calibration plot, and unit normal loss integral in two dimensions.

You can also access the vignettes from R:

browseVignettes("predtools")