Gamma regression with a log-link
Gamma regression with a log-link.
gammareg(y, x, tol = 1e-07, maxiters = 100) gammacon(y, tol = 1e-08, maxiters =50)
y |
The dependent variable, a numerical variable with non negative numbers. |
x |
A matrix or data.frame with the indendent variables. |
tol |
The tolerance value to terminate the Newton-Raphson algorithm. |
maxiters |
The maximum number of iterations that can take place in the regression. |
The gamma.reg fits a Gamma regression with a log-link. The gamma.con fits a Gamma regression with a log link with the intercept only ( glm(y ~ 1, Gamma(log) ) ).
A list including:
deviance |
The deviance value. |
phi |
The dispersion parameter (φ) of the regression. This is necessary if you want to perform an F hypothesis test for the significance of one or more independent variables. |
be |
The regression coefficient(s). |
info |
The number of iterations, the deviance and the dispersion parameter. |
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@yahoo.gr.
McCullagh, Peter, and John A. Nelder. Generalized linear models. CRC press, USA, 2nd edition, 1989.
y <- abs( rnorm(100) ) x <- matrix( rnorm(100 * 2), ncol = 2) mod <- glm(y ~ x, family = Gamma(log) ) res<-summary(mod) ## Not run: res<-gammareg(y, x) ## End(Not run) mod <- glm(y ~ 1, family = Gamma(log) ) res<-summary(mod) res<-gammacon(y)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.