Overview

The calibration routine uses a standard least-squares approach to fit model parameters to observed data. The calibrate function is inherited from WiNDCContainer, but requires specification of exogenous (fixed) variables and constraints to guide the calibration process. Variables are fixed by WiNDCNational.calibrate_fix_variables and the constraints are defined using WiNDCNational.calibrate_constraints.

The calibrate function returns a pair (table, JuMP.Model) where table is a WiNDCTable type and the JuMP.Model is the JuMP model used in the calibration. The JuMP model can be used to determine if the model calibrates and debug the constraints if it does not.