Fit Ensemble of NLS Models for One Plate — fit_ensemble

Fits each model in the ensemble using multi-start Levenberg-Marquardt NLS, returning the best fit per model (by AIC). Handles convergence failures gracefully — a model that fails to converge gets converged = FALSE in the output.

Usage

fit_ensemble_nls(
  data,
  formulas,
  model_constraints,
  start_lists,
  verbose = FALSE
)

Arguments

data: Data frame of preprocessed standards (already transformed by curveRcore::preprocess_standards()).
formulas: Named list of formula objects from curveRcore::build_nls_formulas().
model_constraints: Named list from compute_model_constraints().
start_lists: Named list from generate_start_lists().
verbose: Logical.

Value

A named list (one element per model) of lists with:

model_name: Character.
converged: Logical.
fit: The nlsLM object, or NULL.
aic: Numeric (Inf if failed).
data: The data used for fitting.