modelObjectiveFunction: Objective function to fit model parameters

In pdmod: Proximal/Distal Modeling Framework for Pavlovian Conditioning Phenomena

Description

Function passed to optimization routine to minimize to estimate parameters. Uses mean squared error to calculate difference between dataResponse and what computeModel) would forcast for dataX using parameters pars.

Usage

  modelObjectiveFunction(pars, dimension, dataX,
    dataResponse, responseFunction = calculateResponse,
    sessionBoundaries = NA, fitG = TRUE)

Arguments

pars	Vector of parameters mFast, mSlow, n, hSlow, and r
dimension	What dimension to return error in, 1 for single criteria optimization, or number of columns of data for multicriteria optimization
dataX	List of observations of process x(i) (with real time)
dataResponse	Corresponding list of observations of subject's response to x(i), i.e. ~x(i)
responseFunction	The function to use to transform the forecast into a response
sessionBoundaries	(option) Vector defining how to group the trials into sessions where the items are the starting indicies for each session (so the last value can be the index after the last trial) and NAs are used for gaps between sessions
fitG	TRUE to estimate g, or FALSE to fix g at 0

Value

Error between dataRespones and what would have been estimated for dataX based on parameters pars

Author(s)

Chloe Bracis

See Also

computeModel, fitModel

pdmod documentation built on May 2, 2019, 5:16 a.m.