control_staninput: Specify control parameters for make_staninput()

In hlplab/MVBeliefUpdatr: Fitting, Summarizing, and Visualizing of Gaussian Ideal Observers and Adaptors

Specify control parameters for make_staninput()

Description

This function is used to specify control parameters for the 'make_staninput()' function, and to provide reasonable defaults for any of the unspecified parameters.

Usage

control_staninput(
  tau_scale = 5,
  L_omega_eta = 1,
  split_loglik_per_observation = 0,
  transform_type = c("identity", "center", "standardize", "PCA whiten", "ZCA whiten")[3]
)

Arguments

tau_scale	A vector of scales for the Cauchy priors for each cue's standard deviations. Used in both the prior for m_0 and the prior for S_0. (default: vector of '5's, assuming that the data are standardized).
L_omega_eta	A vector of etas of the LKJ prior for the correlations of the covariance matrix of mu_0. Only used for models with multivariate categories (e.g., NIW_ideal_adaptor). (default: '1', which corresponds to a uniform prior of correlation matrices)
split_loglik_per_observation	Optionally, split the log likelihood per observation. This can be helpful of leave-one-out estimation in order to avoid high Pareto k, but it also makes the stored stanfit object much larger. (default: '0')
transform_type	An affine transformation that can be applied to the data. See 'type' in get_affine_transform for details. (default: "standardize", which standardizes each cue separately)

Value

A list of control parameters that can be passed to make_staninput.

hlplab/MVBeliefUpdatr documentation built on July 5, 2025, 6:42 a.m.