Fitting Additive Binomial Regression Models

Share:

Description

Workhorse function for addreg with binomial family.

Usage

1
addbin(y, x, start = NULL, control = list(), allref)

Arguments

y

binomial response. May be a single column of 0/1 or two columns, giving the number of successes and failures.

x

non-negative design matrix. Must have an intercept column.

start

starting values for the parameters in the linear predictor.

control

list of parameters for controlling the fitting process, passed to addreg.control.

allref

a list of all parameterisations for this model, obtained from addreg.allref.

Details

An additive binomial fit can be converted into an additive Poisson fit via the multinomial–Poisson transformation (Baker, 1994). This function transforms the data as described by Donoghoe and Marschner (2014) and passes it to addreg with a Poisson family to get the maximum likelihood estimate. The coefficients (and other values) from the Poisson model are transformed back to relate to the additive binomial model.

This is a workhorse function for addreg when a binomial family is specified. It would not usually be called directly.

Value

A list of (most of) the components needed for an object of class "addreg"; see addreg for details.

Author(s)

Mark W. Donoghoe Mark.Donoghoe@mq.edu.au

References

Baker, S. G. (1994). The multinomial–Poisson transformation. The Statistician 43(4): 495–504.

Donoghoe, M. W. and I. C. Marschner (2014). Stable computational methods for additive binomial models with application to adjusted risk differences. Computational Statistics and Data Analysis 80: 184–196.

See Also

addreg