Learn R Programming

boostr (version 1.0.0)

wrapReweighter: Create a boostr compatible wrapper for a reweighter.

Description

Use provided metadata on a given reweighter to create a boostr compatible wrapper.

Usage

wrapReweighter(reweighter, reweighterInputPreds = "prediction", reweighterInputResponse = "response", reweighterInputWts = "weights", reweighterOutputWts = "weights", .verbose = FALSE)

Arguments

reweighter
a function which satisfies the abstract definition of a reweighter (see description below).
reweighterInputPreds
a string indicating the name of the argument reweighter uses to represent the input predictions.
reweighterInputResponse
a string indicating the name of the argument reweighter uses to represent the true responses for the input predictions.
reweighterInputWts
a string indicating the name of the argument reweighter uses to represent the input weights.
reweighterOutputWts
a string indicating the name of the entry in reweighter's output that represents the output weights.
.verbose
a boolean indicating if warnings should be displayed or not.

Value

A function (wrapper around reweighter) which is a 'reweighter' object. The wrapper's signature is (prediction, response, weights, ...) and its output is a list that names the cell containing its weights 'weight'. Hence, the wrapper is a boostr compatible reweighter.

Reweighters

See the Reweighters section in the vignette vignette(topic = "boostr_user_inputs", package="boostr") for more details on reweighters.

References

Steven Pollack. (2014). Boost: a practical generalization of AdaBoost (Master's Thesis). http://pollackphoto.net/misc/masters_thesis.pdf

See Also

Other Wrapper Generators: buildEstimationProcedure; wrapAggregator; wrapPerformanceAnalyzer; wrapProcedure