Skip to contents

Rule-based and deliberately conservative: it names the parameters and the reasoning, it does not build or launch a calibration run itself. Meant to be read, not blindly executed - a stage design still has to reconcile these suggestions against every other lake being calibrated in the same run.

Usage

recommend_calib_plan(diag, param_table = NULL)

Arguments

diag

an aeme_diag object from diag_aeme.

param_table

data frame with columns name (parameter name) and vars (list-column; each element a character vector of the AEME var_sim names that parameter is linked to), used to look up which calibration parameters are linked to a biased nutrient-budget variable. NULL (default) skips that lookup.

Value

data frame with columns issue, severity ("low"/"medium"/"high"), evidence, recommendation, params (comma-joined parameter names, where known).