glm.lld {gplm}

Log-likelihood derivatives for GLM

Description

Computes first and second derivatives of the individual log-likelihood with respect to the linear predictor. Currently only the gaussian (with identity link) and the bernoulli family (with logit and probit links) are implemented.

Usage

glm.lld(eta, y, family="gaussian", link="identity", k=1)

Arguments

eta

n x 1, linear predictors

y

n x 1, responses

family

text string, family of distributions (e.g. "gaussian" or "bernoulli", see details for glm.ll)

link

text string, link function (depending on family, see details for glm.ll)

k

integer > 0, parameter for the negative binomial

Details

See details for glm.ll.

Value

List with components:

ll1

n x 1, vector of first derivatives

ll2

n x 1, vector of second derivatives

ll1.2

n x 1, ratio ll1/ll2

Author(s)

Marlene Mueller

See Also

glm.ll, glm.link

Examples

  glm.lld(c(-1,2), c(0,1), family="bernoulli", link="logit")