Sample parameters from an external file

Create parameters for a simulated population by sampling them from an external file

Usage

createExternalParameters(
  subjects,
  file,
  names,
  range,
  betNames,
  betNums,
  errStruc = c("None", "Proportional", "Additive", "Log-Normal"),
  subset = NULL,
  refCol,
  refColName,
  idCol = getEctdColName("Subject"),
  seed = .deriveFromMasterSeed(),
  flagName = getEctdColName("ParOmit"),
  refColSuffix = "refCol",
  dataId = idCol,
  workingPath = getwd()
)

Arguments

subjects: (Required) Number of subjects for which to create parameters
file: (Required) External file from which to extract parameters
names: (Required) Names of parameters to import (refers to variables in the input data)
range: (Optional) Range of acceptable values. Missing by default, resulting in no "range" of values is applied to the data
betNames: (Required) Between subject effects variables in the data
betNums: (Required) Integer vector mapping between subject effects onto fixed effects
errStruc: (Optional) (None) Function to map between subject effects: Additive, Proportional, Log-Normal or None
subset: (Optional) Subset to apply to the data. By default, no subset is applied
refCol: (Optional) Vector of data to be matched to the reference column for sampling (no reference data by default)
refColName: (Optional) Column name in data for referenced sampling (no reference column by default)
idCol: (Optional) Subject variable name in the data ("SUBJ" by default)
seed: (Optional) Random number seed. Derived from the current seed by default
flagName: (Optional) Name for "parameter omit" flag ("PAROMIT" by default)
refColSuffix: (Optional) Suffix to add to reference variable in the data ("refCol" by default)
dataId: (Optional) The subject variable name in the data. By default, this is taken to be the same as the "idCol" input
workingPath: (Optional) Working path from which to import covariate file. By default, the working path will be used

Value

A data frame containing a subject variable, a "parameter omit flag" and a set of extracted parameters

Details

The data is first imported from the specified file, and any subset applied to it. The function next checks that all required variables specifed appear in the imported data. Fixed effects (specified using names) are selected from the file along with between subject effects (if they are also identified).

If fixed and between subject effects have been generated, the "errStruc" input will specify how the between subject effects are to be applied: * None = Keep between subject effect seperate (and use the "suffix" supplied to name the variables) * Additive = Add each between subject effect to the corresponding fixed effect * Proportional = Multiply the fixed effect by the exponentiated between subject effect * Log-Normal = Add "1 + each between subject effect" to the corresponding fixed effect

If the relCol input is specified, this variable should also be retained in the result. The "parameter omit" flag will contain repeated "0" elements in the return result.

Note

If earlier versions of MSToolkit, the "Proportional" error structure was implemented as exp(fixed + between subject)

Author

Mike K Smith mstoolkit@googlemail.com

Examples


if (FALSE) {
  wPath <- system.file( "Runit", "data", "createParameters", package = "MSToolkit" )

  createExternalParameters( subjects = 100, names = "E0,ED50",
    file = "testParam.csv", errStruc = "None", betNames = "B1,B2", workingPath = wPath )
}