logit_link {EstimationTools}

Logit link function (for estimation with maxlogL object)

Description

log_link object provides a way to implement logit link function that maxlogL needs to perform estimation. See documentation for maxlogL for further information on parameter estimation and implementation of link objects.

Usage

logit_link()

Details

logit_link is part of a family of generic functions with no input arguments that defines and returns a list with details of the link function:

1. name: a character string with the name of the link function.

2. g: implementation of the link function as a generic function in R.

3. g_inv: implementation of the inverse link function as a generic function in R.

There is a way to add new mapping functions. The user must specify the details aforesaid.

Value

A list with logit link function, its inverse and its name.

Author(s)

Jaime Mosquera Gutiérrez, [e-mail]

See Also

maxlogL

Other link functions: NegInv_link(), log_link()

Examples

#--------------------------------------------------------------------------------
# Estimation of proportion in binomial distribution with 'logit' function
# 10 trials, probability of success equals to 30%)
N <- rbinom(n = 100, size = 10, prob = 0.3)
phat <- maxlogL(x = N, dist = 'dbinom', fixed = list(size=10),
                link = list(over = "prob", fun = "logit_link"))
summary(phat)

# Link function name
fun <- logit_link()$name
print(fun)

# Link function
g <- logit_link()$g
curve(g(x), from = 0, to = 1)

# Inverse link function
ginv <- logit_link()$g_inv
curve(ginv(x), from = -10, to = 10)

#--------------------------------------------------------------------------------