Fixed intercepts Poisson regression {Rfast2}

Fixed intercepts Poisson regression

Description

Fixed intercepts Poisson regression.

Usage

fipois.reg(y, x, id, tol = 1e-07, maxiters = 100)

Arguments

y

The dependent variable, a numerical vector with integer, non negative valued data.

x

A matrix with the indendent variables.

id

A numerical variable with 1, 2, ... indicating the subject. Unbalanced design is of course welcome.

tol

The tolerance value to terminate the Newton-Raphson algorithm. This is set to 10^{-7} by default.

maxiters

The maximum number of iterations that can take place during the fitting.

Details

Fixed intercepts Poisson regression for clustered count data is fitted. According to Demidenko (2013), when the number of clusters (N) is small and the number of observations per cluster (n_i) is relatively large, say min(n_i) > N, one may assume that the intercept \alpha_i = \beta + u_i is fixed and unknown (i=1,...,N).

Value

A list including:

be

The regression coefficients.

seb

The standard errors of the regression coefficients.

ai

The estimated fixed intercepts fore ach cluster of observations.

covbeta

The covariance matrix of the regression coefficients.

loglik

The maximised log-likelihood value.

iters

The number of iteration the Newton-Raphson required.

Author(s)

Michail Tsagris.

R implementation and documentation: Michail Tsagris [e-mail].

References

Eugene Demidenko (2013). Mixed Models: Theory and Applications with R, pages 388-389, 2nd Edition. New Jersey: Wiley & Sons (excellent book).

See Also

cluster.lm, fe.lmfit, gee.reg, covar, welch.tests

Examples

y <- rpois(200, 10)
id <- sample(1:10, 200, replace = TRUE)
x <- rpois(200, 10)
fipois.reg(y, x, id)