single_model_ll: Top level log-likelihood function that implements the single...
In gjcooper/gcphd-model_of_dce: Models of Consumer Choice Experiments

single_model_ll

R Documentation

Top level log-likelihood function that implements the single model sampling

Description

This function runs one of the possible architectures repeatedly, horrible, uses a global variable for the selected model.

Usage

single_model_ll(
  x,
  data,
  contaminant_prob = 0.02,
  architecture = "IST",
  min_rt = 0,
  max_rt = 1
)

single_model_ll(
  x,
  data,
  contaminant_prob = 0.02,
  architecture = "IST",
  min_rt = 0,
  max_rt = 1
)

Arguments

`x`	A named vector containing parameter values to test
`data`	The data for a single subject for which the likelihood should be calculated
`contaminant_prob`	The probability used for contaminant process in the modelling. A contaminant process is just a uniform random response in the allowable time window.
`architecture`	The name of the architecture model to implement
`min_rt`	The smallest possible response time in the data
`max_rt`	The largest possible response time in the data

Value

The log of the likelihood for the data under parameter values x

The parameter vector

The vector x should contain the following elements: A number of \alpha values

A - the start point variability
b^a and b^r, the thresholds to either accept or reject the item.
t0 - the residual time, bounded above by the minimum response time for the participant
12 drift rates. For each attribute there are three stimulus levels. For each of these 6 attribute levels there are two drift rates, one drift rate to accept (v^a) and one to reject (v^r)

The vector x should contain the following elements: A number of \alpha values

A - the start point variability
b^a and b^r, the thresholds to either accept or reject the item.
t0 - the residual time, bounded above by the minimum response time for the participant
12 drift rates. For each attribute there are three stimulus levels. For each of these 6 attribute levels there are two drift rates, one drift rate to accept (v^a) and one to reject (v^r)