# gldrmPIT: Confidence intervals for gldrm coefficients In gldrm: Generalized Linear Density Ratio Models

## Description

Plots and returns the randomized probability inverse transform of a fitted gldrm.

## Usage

 1 2 gldrmPIT(gldrmFit, nbreaks = 7, cex.main = NULL, cex.lab = NULL, cex.axis = NULL) 

## Arguments

 gldrmFit A gldrm model fit. Must be an S3 object of class "gldrm", returned from the gldrm function. The matrix of semiparametric tilted probabilities must be returned, which is done by fitting gldrm with gldrmControl = gldrm.control(returnfTiltMatrix = TRUE). nbreaks Number of breaks in the histogram. cex.main Text size for main titles. cex.lab Text size for axis labels. cex.axis Text size for axis numbers.

## Details

The probability inverse transform is defined generally as \hat{F}(y|x), which is the fitted conditional cdf of each observation evaluated at the observed response value. In the case of gldrm, the fitted cdf is descrete, so we draw a random value from a uniform distribution on the interval (\hat{F}(y|x), \hat{F}(y-|x)), where y- is the next largest observed support less than y (or -Infinity if y is the minimum support value). The output and plots generated by this function will vary slightly each time it is called (unless the random number generator seed is set beforehand).

## Value

Randomized robability inverse transform as a vector. Also plots the histogram and uniform QQ plot.

## Examples

 1 2 3 4 5 6 7 8 data(iris, package="datasets") ### Fit gldrm and return fTiltMatrix fit <- gldrm(Sepal.Length ~ Sepal.Width + Petal.Length + Petal.Width + Species, data=iris, link="log") # Probability inverse transform plot gldrmPIT(fit) 

gldrm documentation built on May 2, 2019, 12:59 p.m.