% \VignetteIndexEntry{Vaso Constriction - Logistic Regression}
%\VignetteEngine{knitr::knitr}
%\VignetteEncoding{UTF-8}
\documentclass[a4paper]{article}

\title{Vaso Constriction - Logistic Regression}

\begin{document}

\maketitle
First the dataset vaso is loaded.
<<results='hide',eval=TRUE>>=
library(catdata)
data(vaso)
attach(vaso)
@
For the fitting of a logit model, the response is 0-1 coded. (data set contains 1 2).
Moreover, the covariates vol and rate are log-transformed.
<<eval=TRUE>>=
y <- vaso$vaso
y[vaso$vaso==2] <- 0
@
Fit of a logit-model with log-transformed covariates.
<<eval=TRUE>>=
vaso1 <- glm(y ~ vol + rate, family=binomial)
summary(vaso1)
@
Next,  a logit-model with original covariates is fitted.
<<eval=TRUE>>=
vaso2 <- glm(y ~ I(exp(vol)) + I(exp(rate)), family=binomial)
summary(vaso2)
@
\end{document}