S.fitFH: Model Selection Fay Herriot based on the CME

In stalynGuerrero/multisae: Small Area Estimation using plaucible values

Description

Function that selects a Fay Herriot model based on the mean squared error (MSE), based on all possible combinations of p variables

Usage

S.fitFH(x, yhat, Sd.yhat, Xk, p = 1, y = NULL)

Arguments

x	data.frame which contains the model´s variables.
yhat	Name of the column in x which indicates the variable to estimate.
Sd.yhat	Name of the column in x which indicates the standard deviation estimated for yhat.
Xk	Vector which contains the names of the auxiliary variables. These must be contained in x.
p	Maximum number of combinations to make with the covariantes. By default p = 1.
y	Optional. Name of the column in x that has the actual values of the estimated variable.

Value

Returns an object of the data.frame class, which has the summary measures employed for the selection of the desired model.

model

List of models resulting from ∑_{i=1}^{p} {{p}\choose{i}}

loglike

Quality measures related to the model based on the likelihood function

mse.Sample

Mean squared error obtained for observations without missing data in their covariantes.

N.sample

Number of observations without missing data.

mse.predict

Mean squared error obtained for the forecasts.

N.predict

Number of observations with prediction

mse.predictIMP

Mean square error obtained for predictions with observations missing covariates

N.predictIMP

Number of observations with missing data

References

-

Fay, R.E. and Herriot, R.A. (1979). Estimation of income from small places: An application of James-Stein procedures to census data. Journal of the American Statistical Association 74, 269-277.

-

Marhuenda, Y., Morales, D. and Pardo, M.C. (2014). Information criteria for Fay-Herriot model selection. Computational Statistics and Data Analysis 70, 268-280

-

Rao, J.N.K. (2003). Small Area Estimation. Wiley, London.

Examples

data('etc2013')
Xk <- c('x1','x2','x3','x4','x5','x6','x7','x8','x9')
Sel.fit <- S.fitFH(etc2013, 'hat.prom', 'hat.sd',Xk, p = 2)
Sel.fit <- S.fitFH(etc2013, 'hat.prom', 'hat.sd',Xk, p = 2, y = 'prom')

stalynGuerrero/multisae documentation built on May 30, 2019, 8:44 a.m.