Construct a symbolized_model object — new_symbolized

new_symbolized_model() is the internal S3 constructor. It validates field presence and types and returns an object of class "symbolized_model". Tier-specific extractors such as symbolize.drmTMB() call this constructor to wrap the fields they have populated.

Usage

new_symbolized_model(
  model,
  index,
  parameterization,
  distribution,
  submodels,
  terms,
  fixed_effects,
  random_effects = NULL,
  variance_components = NULL,
  covariance_components = NULL,
  loadings = NULL,
  factor_coding = NULL,
  symbol_dictionary,
  assumptions,
  components,
  interpretation,
  formula_bridge,
  warnings_registry = NULL,
  graph = NULL,
  expanded = NULL,
  metadata
)

Arguments

model: A list with at least class, package, family, response, n_obs.
index: A list of index symbols (observation, individual, group, trait, time).
parameterization: A list capturing the family-specific scale meaning.
distribution: A tibble of response distribution rows.
submodels: A tibble with one row per linked distributional parameter.
terms: A tibble: the term-grammar / model-matrix bridge.
fixed_effects: A tibble of fixed-effect estimates joined to terms.
random_effects: A tibble or NULL.
variance_components: A tibble or NULL.
covariance_components: A tibble or NULL.
loadings: A tibble of latent-factor loadings (rows: one per (submodel, trait, axis) entry of a loading matrix) or NULL. Populated by extractors for reduced-rank latent-variable models such as gllvmTMB.
factor_coding: A tibble (one row per factor predictor) recording each factor's levels, reference_level, contrast_type, n_dummies, and is_default_treatment, or NULL. Populated during extraction by build_factor_coding().
symbol_dictionary: A tibble of (symbol, variable, units, role, description).
assumptions: A tibble of stated/implied assumptions.
components: A tibble: one row per renderable block.
interpretation: A tibble of per-parameter readings.
formula_bridge: A tibble: R syntax to statistical meaning to mathematics.
warnings_registry: A tibble or NULL.
graph: A list or NULL.
expanded: A list of actual numeric arrays (response vector, design matrices, coefficient vectors, random-effect BLUPs, residuals, fitted mu/sigma) or NULL. Populated by extractors that have access to the original fit; consumed by expand() and the three-views renderer.
metadata: A list with at least call, context, package_versions, created_by.

Value

A symbolized_model S3 object.