Create Study-Level Fitting Parameters — new_study

Constructs the study-wide settings that control data preprocessing.

Usage

new_study_params(
  is_log_response = TRUE,
  is_log_independent = TRUE,
  apply_prozone = TRUE,
  blank_option = "ignored",
  persist_draws = FALSE,
  bayes_single_plate = FALSE
)

Arguments

is_log_response: Logical. Log10-transform the assay response? Default TRUE.
is_log_independent: Logical. Log10-transform the concentration? Default TRUE.
apply_prozone: Logical. Apply prozone (hook effect) correction? Default TRUE.
blank_option: Character. Blank handling method. One of "ignored", "included", "subtracted", "subtracted_3x", "subtracted_10x". Default "ignored".
persist_draws: Logical. If TRUE, the fitting engine persists the full posterior (Bayesian) or asymptotic-MVN (frequentist) parameter draws to calib_draws. Heavy output; default FALSE. Gate only — the light calib_hyperparam/calib_fit_diag tables are always written.
bayes_single_plate: Logical. Bayesian only. If TRUE, each plate is fit independently (N_plates = 1, no cross-plate pooling) instead of as one multiplate hierarchy. Default FALSE (multiplate pooling). The frequentist engine is always per-plate and ignores this flag.

Value

A named list of class study_params.