summarize_indicators: Presents point, MSE and CV estimates

In krennpa/SAEforest: Mixed Effect Random Forests for Small Area Estimation

Presents point, MSE and CV estimates

Description

Function summarize_indicators reports point and mean squared error (MSE) estimates as well as calculated coefficients of variation (CV) from a fitted SAEforest object.

Usage

summarize_indicators(object, indicator = "all", MSE = FALSE, CV = FALSE)

Arguments

object	Object for which point and/or MSE estimates and/or calculated CV's are requested. The object must be of class SAEforest.
indicator	Optional character vector specifying indicators to be mapped: (i) all calculated indicators ("all"); (ii) each default indicators name: "Mean", "Quant10", "Quant25", "Median", "Quant75", "Quant90", "Gini", "Hcr", "Pgap", "Qsr" or the function name/s of "custom_indicator/s"; (iii) a vector of names of indicators. If the object is estimated by SAEforest_model indicator arguments are ignored and only the "Mean" is returned.
MSE	Logical. If TRUE, MSE estimates for selected indicators per domain are added to the data frame of point estimates. Defaults to FALSE.
CV	Logical. If TRUE, coefficients of variation for selected indicators per domain are added to the data frame of point estimates. Defaults to FALSE.

Details

Objects of class summarize_indicators.SAEforest have methods for following generic functions: head and tail (for default documentation, see head), as.matrix (for default documentation, see matrix), as.data.frame (for default documentation, see as.data.frame), subset (for default documentation, see subset).

Value

The return of summarize_indicators is an object of class summarize_indicators.SAEforest including domain-specific point and/or MSE estimates and/or calculated CV's from a SAEforest object The returned object contains the data.frame ind and a character including the names of requested indicator(s).

See Also

SAEforestObject, SAEforest_model

Examples

# Loading data
data("eusilcA_pop")
data("eusilcA_smp")

income <- eusilcA_smp$eqIncome
X_covar <- eusilcA_smp[, -c(1, 16, 17, 18)]

# Calculating point + MSE estimates and passing arguments to the forest.
# Additionally, two additional indicators and functions as threshold are added.
# Note that B and num.trees are low to speed up estimation time and must be changed for
# practical applications.

model1 <- SAEforest_model(Y = income, X = X_covar, dName = "district",
                          smp_data = eusilcA_smp, pop_data = eusilcA_pop,
                          meanOnly = FALSE, MSE = "nonparametric", B = 5, mtry = 5,
                          num.trees = 50, smearing = FALSE)

# Extract indicator and show generics:
Gini1 <- summarize_indicators(model1, MSE = TRUE, CV = TRUE, indicator = "Gini")

head(Gini1)
tail(Gini1)
as.data.frame(Gini1)
as.matrix(Gini1)
subset(Gini1, district == "Wien")

krennpa/SAEforest documentation built on Jan. 30, 2024, 12:31 a.m.