get_starting_values: Get Starting Values for Model Parameters

In dRiftDM: Estimating (Time-Dependent) Drift Diffusion Models

Get Starting Values for Model Parameters

Description

I return a named numeric vector of starting values, when possible. If use_ez is TRUE and obs_data is available, I try to obtain EZ-diffusion-based guesses for compatible component functions. Remaining parameters (if any) can be initialized via a simple latin hypercube search within lower/upper.

Usage

get_starting_values(object, ...)

## S3 method for class 'drift_dm'
get_starting_values(
  object,
  ...,
  lower = NULL,
  upper = NULL,
  verbose = 0,
  use_ez = NULL,
  n_lhs = NULL
)

Arguments

object	supported object, currently only of type drift_dm.
...	additional arguments passed to the respective method.
lower, upper	optional named numeric vectors of search bounds. Required for the latin hypercube search.
verbose	numeric scalar controlling messages (0 = silent).
use_ez	logical; try EZ-diffusion starting values when possible.
n_lhs	integer; base size factor for latin hypercube sampling (total samples are n_lhs * d, where d is the number of remaining parameters).

Value

a named numeric vector of starting values, or NULL if no reasonable starting values can be determined.

dRiftDM documentation built on Aug. 23, 2026, 5:11 p.m.