buildEnsembleModel: Build Ensemble Model

Description Usage Arguments Details Value

View source: R/buildEnsembleModel.R


This function should build an R object which contains all the information from the ensemble models. Thus, the object should be able to be loaded and used to fill in imputed values and generate predictions using the originally developed model (rather than requiring the model be refit each time).


buildEnsembleModel(data, imputationParameters, processingParameters)



The data.table object containing the dataset we wish to impute on.


A list of the parameters for the yield imputation. See defaultImputationParameters() for a starting point.


A list of the parameters for the production processing algorithms. See defaultProductionParameters() for a starting point.


To actually construct and save such a model object is unfortunately quite a large task: we must save a model object for each individual model in addition to the weights/errors from fitting these models to the data. Thus, instead of this task, we can just save the imputations and fill those in when the impute module is called. This also gives us the ability to version control the imputations: we could put them in a new table in the SWS.


A list containing the model fit, model errors, and model weights.

SWS-Methodology/faoswsImputation documentation built on April 7, 2018, 10:12 p.m.