addNoDataModels: Add Ensemble Weights for Sets of the Data without Valid...

In SWS-Methodology/faoswsImputation: Package to perform ensemble imputation for various FAO domains

Description

In some cases, countries (or whatever the byKey is) will not have any data available. In these cases, we can impute values for these observations via an ensemble, but we'll need a method for weighting the models in the ensemble.

Usage

addNoDataModels(data, weights, imputationParameters)

Arguments

data	The data.table object containing the data which will be imputed.
weights	A data.table containing the currently estimated ensemble weights and errors. These are used to compute the global errors and add new weights for models with missing data.
imputationParameters	A list of the parameters for the imputation algorithms. See defaultImputationParameters() for a starting point.

Details

This function computes the weights for these models by looking at the cross-validation error of the global models and uses those errors to determine the final weights.

Value

A data.table of the same format as weights. This function is really just a helper function for getEnsembleWeight, and so the output of the function is specifically designed to be rbinded on to the weights object in that function.

SWS-Methodology/faoswsImputation documentation built on May 9, 2019, 11:48 a.m.