ScenarioInfo: ScenarioInfo constructor

In JGCRI/gcamland: GCAM Land Allocation Module

Description

Create a structure that ontains all information needed to describe a scenario.

Usage

ScenarioInfo(
  aExpectationType = NULL,
  aLaggedShareOld1 = NA,
  aLaggedShareOld2 = NA,
  aLaggedShareOld3 = NA,
  aLinearYears1 = NA,
  aLinearYears2 = NA,
  aLinearYears3 = NA,
  aLogitUseDefault = TRUE,
  aLogitAgroForest = NA,
  aLogitAgroForest_NonPasture = NA,
  aLogitCropland = NA,
  aUseZeroCost = FALSE,
  aCalibrateShareWt = TRUE,
  aIncludeSubsidies = FALSE,
  aShareWeights = NULL,
  aScenarioType = DEFAULT.SCENARIO.TYPE,
  aScenarioName = NULL,
  aFileName = NULL,
  aOutputDir = "./outputs",
  aSerialNum = NA,
  aRegion = DEFAULT.REGION,
  aSubRegion = NULL
)

Arguments

aExpectationType	Expectation type
aLaggedShareOld1	Share of old expectations included in current expectation for crop group 1 (see constants.R)
aLaggedShareOld2	Share of old expectations included in current expectation for crop group 2 (see constants.R)
aLaggedShareOld3	Share of old expectations included in current expectation for crop group 3 (see constants.R)
aLinearYears1	Years for linear expectations for crop group 1 (see constants.R)
aLinearYears2	Years for linear expectations for crop group 2 (see constants.R)
aLinearYears3	Years for linear expectations for crop group 3 (see constants.R)
aLogitUseDefault	Boolean indicating whether to use default logits
aLogitAgroForest	AgroForest logit exponent (assuming mLogitUseDefault == FALSE)
aLogitAgroForest_NonPasture	AgroForest_NonPasture logit exponent (assuming mLogitUseDefault == FALSE)
aLogitCropland	Cropland logit exponent (assuming mLogitUseDefault == FALSE)
aUseZeroCost	Boolean indicating whether to set costs to zero (assuming mUseZeroCost == FALSE)
aCalibrateShareWt	Boolean indicating that the model should calculate share weights during calibration
aIncludeSubsidies	Boolean indicating whether subsidy information should modify profit
aShareWeights	Named vector of share weights to use instead of calibrating. The names should correspond to the names of the land leaf nodes.
aScenarioType	Type of scenario to run: either "Reference" or "Hindcast".
aScenarioName	Complete scenario name, with expectations & logit info
aFileName	File name
aOutputDir	Output directory
aSerialNum	Serial number for a run that is part of a series.
aRegion	Region to use in the calculation. Right now we only run a single region at a time.
aSubRegion	Subregion name. Note we can only run a full region or a subregion not both

Details

Most of the parameters are self-explanatory. The Obsvar parameter is a little unusual, in that it plays no role in the model calculations; it only enters into the likelihood function. Fitting this parameter allows us to estimate how much of the variation in the observed data isn't captured by our model. This variation could be because of irreducible uncertainty (e.g., measurement error in the observed data), or it could be an indicator that there is some behavior that our model is failing to capture.

Value

New ScenarioInfo object

Author(s)

KVC November 2017

JGCRI/gcamland documentation built on Oct. 6, 2020, 5:30 p.m.