Quasi Poisson regression for count data {Rfast}

Quasi Poisson regression

Description

Quasi Poisson regression.

Usage

qpois.reg(x, y, full = FALSE, tol = 1e-09,maxiters = 100) 
qpois.regs(x, y, tol = 1e-09, logged = FALSE) 

Arguments

x

For the "qpois.reg" a matrix with data, the predictor variables. This can be a matrix or a data frame. For the "qpois.regs" this must be a numerical matrix, where each columns denotes a variable.

y

A numerical vector with positive discrete data.

full

If this is FALSE, the coefficients, the deviance and the estimated phi parameter will be returned only. If this is TRUE, more information is returned.

tol

The tolerance value to terminate the Newton-Raphson algorithm. This is set to 10^{-9} by default.

logged

Should the p-values be returned (FALSE) or their logarithm (TRUE)?

maxiters

The maximum number of iterations before the Newton-Raphson is terminated automatically.

Details

We are using the Newton-Raphson, but unlike R's built-in function "glm" we do no checks and no extra calculations, or whatever. Simply the model, unless the user requests for the Wald tests of the coefficients. The "qpois.regs" is to be used for very many univariate regressions. The "x" is a matrix in this case and the significance of each variable (column of the matrix) is tested.

Value

For the "prop.reg" a list including: When full is FALSE

be

The regression coefficients.

devi

The deviance of the model.

varb

The covariance matrix of the beta coefficients.

phi

The phi parameter, the estimate of dispersion.

When full is TRUE, the additional item is:

info

The regression coefficients, their standard error, their Wald test statistic and their p-value.

For the "prop.regs" a two-column matrix with the test statistics (Wald statistic) and the associated p-values (or their loggarithm).

Author(s)

Manos Papadakis and Marios Dimitriadis

R implementation and documentation: Manos Papadakis <[e-mail]> and Marios Dimitriadis <[e-mail]>.

References

McCullagh, Peter, and John A. Nelder. Generalized linear models. CRC press, USA, 2nd edition, 1989.

See Also

prop.reg univglms, score.glms, poisson_only

Examples

y <- rnbinom(100, 10, 0.6)
x <- matrix(rnorm(100*3), ncol = 3)
mod1 <- glm(y ~ x, quasipoisson)
res<-summary(mod1)
res<-qpois.reg(x, y, full = TRUE)
res<-qpois.regs(x, y)