diagBinom: Plot Diagnostics for Logistic Regressions

In rmngb: Miscellaneous Collection of Functions for Medical Data Analysis

Description

Two plots are available: a plot of predicted against observed values, and an ROC curve assessing the classifying capacity of the model.

Usage

diagBinom(mod, type = 1:2,
ask = prod(par("mfcol")) < length(type) && dev.interactive())

Arguments

mod	a glm with a binomial family.
type	the plots wanted, with 1 for predicted against observed and 2 for an ROC curve.
ask	logical, if TRUE the user is asked before continuing to a new plot.

Details

More plots may be available in the future.

Value

Nothing of interest.

Author(s)

Antoine Filipovic Pierucci

References

Inspired by the book by Hosmer and Lemeshow.

See Also

plot.lm.

Examples

require(stats)
model1 <- glm(case ~ spontaneous+induced, data = infert, family = binomial())
diagBinom(model1)

rmngb documentation built on May 29, 2017, 9:22 p.m.