Display Hosmer-Lemeshow statistic and table of probabilities following logistic regression using glm with binomial family.

Description

Provides a Hosmer-Lemeshow statistic and table following logistic regression.

Usage

HLTest(obj, g)

Arguments

Format

x

The function has two arguments: model name, number of groups

Details

HLTest is a post-estimation function for logistic regression, following the use of glm(). Usage displays a table of observed vs predicted groups and an overall H-L goodness-of-fit statistic.

Value

list

Note

HLTest must be loaded into memory in order to be effectve. As a function in LOGIT, it is immediately available to a user. My thanks to Bilger and Loughlin for the use of their function.

Author(s)

Adapted from Loughlin, T.M. in Bilder and Loughlin, 2015

References

Hilbe, J. M. (2015), Practical Guide to Logistic Regression, Chapman & Hall/CRC.

Bilder, C.R. and Loughlin, T.M. (2015), Analysis of Categorical Data with R, Chapman & Hall/CRC.

Hilbe, J. M. (2009), Logistic Regression Models, Chapman & Hall/CRC.

Hosmer, D.W., Lemeshow, S, and Sturdivant, R.X (2013), Applied Logistic Regression, 3rd ed, Wiley.

See Also

glm

Examples

library(MASS)
library(LOGIT)
data(medpar)
mylogit <- glm( died ~  los + white + hmo, family=binomial, data=medpar)
grp10 <- HLTest(obj=mylogit, g=10)
cbind(grp10$observed, round(grp10$expect, digits = 1))
grp10

Results

R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(LOGIT)
> png(filename="/home/ddbj/snapshot/RGM3/R_CC/result/LOGIT/HLTest.Rd_%03d_medium.png", width=480, height=480)
> ### Name: HLTest
> ### Title: Display Hosmer-Lemeshow statistic and table of probabilities
> ###   following logistic regression using glm with binomial family.
> ### Aliases: HLTest
> ### Keywords: models
> 
> ### ** Examples
> 
> library(MASS)
> library(LOGIT)
> data(medpar)
> mylogit <- glm( died ~  los + white + hmo, family=binomial, data=medpar)
> grp10 <- HLTest(obj=mylogit, g=10)
> cbind(grp10$observed, round(grp10$expect, digits = 1))
                Y0 Y1 Y0hat Y1hat
[0.0213,0.278] 101 52 117.7  35.3
(0.278,0.309]  104 42 102.9  43.1
(0.309,0.325]  112 45 107.2  49.8
(0.325,0.339]  109 37  97.2  48.8
(0.339,0.352]  119 42 105.0  56.0
(0.352,0.366]  137 54 121.8  69.2
(0.366,0.373]   20  3  14.5   8.5
(0.373,0.387]  199 77 171.0 105.0
(0.387,0.402]   61 81  85.5  56.5
(0.402,0.409]   20 80  59.1  40.9
> grp10

	Hosmer and Lemeshow goodness-of-fit test with 10 bins

data:  mylogit
X2 = 124.38, df = 8, p-value < 2.2e-16

> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>