mleLBH: MLE model parameter estimator

In XiaodanLyu/saezero: Small Area Estimation under a Zero Inflated Lognormal Model with Correlated Random Area Effects

Description

Fits by ML method the unit level model of Lyu, Berg and Hofmann. The specified link function is used in the binary part.

Usage

mleLBH(data_2p, link = "logit")

Arguments

data_2p	a two-part data object returned by as.2pdata.
link	a specification for the link function used to model the binary part. The accepted link functions are logit, probit, cauchit, log, cloglog and an object of the class link-glm, see the documentation of make.link. Default value is "logit".

Value

The function returns a list with the following objects:

• fixed: list with the estimated values of the fixed regression coefficient in the positive part (p1) and in the binary part (p0).

• random: data frame with the predicted random effects in the positive part (p1) and the binary part (p0).

• errorvar: estimated model error variance in the positive part.

• refvar1: estimated random effects variance in the positive part.

• refvar0: estimated random effects variance in the binary part.

• refcor: estimated correlation coefficient of the random effects between the two parts.

• loglik: log-likelihood accomodating a general transformation family and lambda in the positive part as specified in as.2pdata.

• residuals: the marginal (mar) and conditional (con) residuals from the model fit in the positive part (p1, cases with nonpositive response are NA.).

• fit0: model parameter estimator under independence assumption.

See Also

as.2pdata

Examples

  erosion_2p <- as.2pdata(f_pos = RUSLE2~logR+logK+logS,
                          f_zero = ~logR+logS+crop2+crop3,
                          f_area = ~cty, data = erosion)
  fit <- mleLBH(erosion_2p)

XiaodanLyu/saezero documentation built on Sept. 20, 2020, 5:59 a.m.