Calibration#
Scientific side of the HydroModPy calibration stack. Complements the architecture pages by answering what is solved, how objectives are built, and how each implemented method behaves numerically.
The scope here is the calibration code in hydromodpy.calibration,
exposed through hmp run <calibration.toml>.
Topic pages#
What inverse problem HydroModPy solves: parameter identifiability, objective construction, prior assumptions, and the difference between a best-fit point and a posterior distribution.
What each built-in optimizer or sampler actually does:
grid, optuna (TPE, CMA-ES), scipy_de,
scipy_nelder_mead, and the distribution-valued methods.
See also#
Calibration Workflow for the operational walkthrough and the “Pick a method” decision table.
Calibration Architecture for the package layout and runtime classes behind the engine.
Calibration Benchmarks for stable benchmark pages with posteriors and objective landscapes.