dmi-class: An S4 Class Representing a Data-Model Instance

In ggdmcModel: Model Builders for 'ggdmc' Package

An S4 Class Representing a Data-Model Instance

Description

The Data-Model Instance, 'dmi', class binds a model specification object with a corresponding experimental dataset. This structure provides a unified container used for fitting cognitive models, simulating responses, or conducting posterior predictive checks.

Details

Unlike the previous version that used data frames, the 'dmi' class expects the data input as a **list** of trial-level records, optimised for internal modelling functions.

Value

An object of class 'dmi' to be passed to functions for model fitting, likelihood evaluation, simulation, or diagnostics.

Slots

model

An object of class 'model' that defines the structure of the cognitive model, including parameter mappings, accumulators, and condition associations.

data

A list representing the observed dataset. Each element typically corresponds to a condition or trial grouping, containing relevant variables (e.g., RTs, responses).

node_1_index

An internal integer matrix used in LBA-type models to indicate the accumulator or node associated with the correct response (e.g., index of node 1).

is_positive_drift

A logical flag used in models where drift direction matters, indicating whether the modelled drift rate is constrained to be positive (in the LBA model) or is going to the positive direction (in the DDM). This can be important for proper interpretation of parameters.

Structure

An object of class 'dmi' has the following slots:

Purpose

This class provides a complete representation of the modelling context by combining the experimental data and model structure. It serves as the standard input to fitting algorithms, allowing for parameter estimation, simulation, and model checking in accumulator-based cognitive models (e.g., LBA, DDM).

ggdmcModel documentation built on Aug. 8, 2025, 7:50 p.m.