emulator_predictions: Used to generate predictions from an emulator, normalising...

In spartan: Simulation Parameter Analysis R Toolkit ApplicatioN: 'spartan'

Description

With an emulator object produced, this can be used to generate predictions on unseen data. This method is called with the emulation object, parameters, meaasures, and unseen data. A flag should also be set as to whether the unseen data, and thus the generated prediction, need to be normalised and rescaled accordingly. Unseen data being input into the emulator must be scaled between 0 and 1, with predictions rescaled after generation.

Usage

emulator_predictions(emulation, parameters, measures, data_to_predict,
  normalise = FALSE, normalise_result = FALSE)

Arguments

emulation	The emulation object to use to make the predictions
parameters	Parameters on which the model will take as input
measures	Simulation responses the model should predict
data_to_predict	Unseen values for the parameters for which the measures should be predicted
normalise	Whether the data_to_predict should be normalised
normalise_result	Whether the resultant predictions should be normalised

Value

Predictions generated for this unseen data

spartan documentation built on May 2, 2019, 9:39 a.m.