Build NLS Formulas for Candidate Models — build_nls

Constructs stats::nls()-compatible formula objects for the requested models, using the curveRcore parameterisation (b > 0, always increasing).

Usage

build_nls_formulas(
  model_names,
  response_variable,
  independent_variable = "concentration",
  fixed_a = NULL,
  is_log_response = TRUE
)

Arguments

model_names: Character vector. Which models to build formulas for. Must be a subset of available_models().
response_variable: Character. Name of the response column in the fitting data frame.
independent_variable: Character. Name of the concentration column in the fitting data frame. Default "concentration".
fixed_a: Numeric scalar or NULL. If non-NULL, the lower asymptote is fixed at this value (on the raw scale before any log transform).
is_log_response: Logical. If TRUE and fixed_a is non-NULL, log10-transform the fixed value before substitution.

Value

Named list of formula objects, one per model.

Details

When fixed_a is non-NULL and is_log_response = TRUE, the raw value is validated and log10-transformed before substitution. When fixed_a is non-NULL and is_log_response = FALSE, it is substituted directly.