glmmML {glmmML} | R Documentation |
Fits GLMs with random intercept by Maximum Likelihood and numerical integration via Gauss-Hermite quadrature.
glmmML(formula, data = list(), cluster = NULL, family = binomial, start.coef = NULL, start.sigma = NULL, offset = NULL, method = "vmmin", control = glm.control(epsilon = 1e-08, maxit = 100, trace = FALSE), n.points = 16)
formula |
a symbolic description of the model to be fit. The details of model specification are given below. |
data |
an optional data frame containing the variables in the model. By default the variables are taken from `environment(formula)', typically the environment from which `glmmML' is called. |
cluster |
Factor indicating which items are correlated |
family |
Currently, the only valid values are binomial and
poisson . The binomial family allows for the logit and
cloglog links, but can only be represented as binary data. |
start.coef |
starting values for the parameters in the linear predictor. Defaults to zero. |
start.sigma |
starting value for the mixing standard deviation. Defaults to 0.5. |
offset |
this can be used to specify an a priori known component to be included in the linear predictor during fitting. |
method |
the method to be used in fitting the model. The default (and presently only) method `vmmin' uses the BFGS method in the 'optim' function. |
control |
Controls the convergence criteria. See
glm.control for details. |
n.points |
Number of points in the Gauss-hermite quadrature. If n.points == 1, an ordinary glm is fitted. |
After the 'vmmin' function has converged, an ordinary Newton-Raphson procedure finishes the maximization. As a by-product, the variance-covariance is estimated.
The return value is a list, an object of class 'glmmML'.
This is a very preliminary version. The optimization may fail with
the default value of start.sigma
. In that case, try different
start values for sigma.
Göran Broström
Broström (2003). Generalized linear models with random intercepts. http://www.stat.umu.se/forskning/glmmML.pdf
optim
, glmm
in Lindsey's
repeated
package, and glmmPQL
in MASS
.
x <- cbind(rep(1, 14), rnorm(14)) y <- rbinom(14, prob = 0.5, size = 1) id <- rep(1:7, 2) dat <- data.frame(y = y, x1 = x[, 2]) glmmML(y ~ x1, data = dat, cluster = id)