Compute curvature-based (shape) LOQs from an enriched grid — compute_shape_loq_from

Identifies the lower shape-LOQ as the concentration at the global maximum of d2y_dx2 and the upper shape-LOQ as the concentration at the global minimum. Vertex positions are refined by 3-point parabolic interpolation.

Usage

compute_shape_loq_from_grid(grid)

Arguments

grid: Data frame with columns log10_concentration, predicted_response, and d2y_dx2.

Value

A named list with elements:

shape_lloq_log10: log10 concentration at the global maximum of d2y_dx2 (lower shape-LOQ).
shape_uloq_log10: log10 concentration at the global minimum of d2y_dx2 (upper shape-LOQ).
shape_lloq_conc: Natural-scale lower shape-LOQ.
shape_uloq_conc: Natural-scale upper shape-LOQ.
shape_lloq_response: Predicted response at shape-LLOQ.
shape_uloq_response: Predicted response at shape-ULOQ.

Details

Designed to be called inside assess_model_eligibility() at Step 3, returning a named list that is merged into the $eligibility slot.