mer function

Minimum Estimated Risk (MER) M-Estimator

mer is an adaptive M-estimator of the weighted mean or total. It is defined as the estimator that minimizes the estimated mean square error, mse, of the estimator under consideration.

mer(object, verbose = TRUE, max_k = 10, init = 1, method = "Brent",
    optim_args = list())

Arguments

• object: an object of class svystat_rob.
• verbose: [logical] indicating whether additional information is printed to the console (default: TRUE).
• init: [numeric] determines the left boundary value of the search interval and the initial value of the search; we must have init < max_k.
• method: [character] the method of optim to be used.
• max_k: [numeric vector] defines the right boundary value of the search interval (default: max_k = 1000)
• optim_args: [list]: arguments passed on to optim.

Details

Package survey must be attached to the search path in order to use the functions (see library or require).

MER-estimators are available for the methods svymean_huber, svytotal_huber, svymean_tukey and svytotal_tukey.

Returns

Object of class svystat_rob

References

Hulliger, B. (1995). Outlier Robust Horvitz-Thompson Estimators. Survey Methodology 21 , 79--87.

See Also

Overview (of all implemented functions)

Examples

head(losdata)

library(survey)
# Survey design for simple random sampling without replacement
dn <- if (packageVersion("survey") >= "4.2") {
        # survey design with pre-calibrated weights
        svydesign(ids = ~1, fpc = ~fpc, weights = ~weight, data = losdata,
                  calibrate.formula = ~1)
    } else {
        # legacy mode
        svydesign(ids = ~1, fpc = ~fpc, weights = ~weight, data = losdata)
    }

# M-estimator of the total with tuning constant k = 8
m <- svymean_huber(~los, dn, type = "rhj", k = 8)

# MER estimator
mer(m)