ABSTRACT

This chapter aims at two extensions: logistic-regression models and nonlinear models. The logistic-regression model covers the situation where the response variable is a binary variable. Logistic regression, which is a particular case of a generalized linear model, arises in several areas, including, for example, analyzing survey data. Logistic regression is an example of a generalized linear model. Nonlinear models can be fit in R using nls. The nls function computes nonlinear least squares. Its usage appears similar to that of lm, but there are differences. The logistic-growth function is identical to that used in logistic regression, although it is written differently. For logistic regression the argument is specified by family=“binomial”, as the default link function. For comparison to simple linear regression, the link function is just an identity, and the family is specified as family=“gaussi".