NestedNewtonThomas and NestedNewtonGC currently lives in it.geoframe.blogspot.numerical.newtonalgorithm, but it directly depends on it.geoframe.blogspot.closureequation.equationstate.EquationState.
Because of this dependency, the numerical module is no longer purely generic/numerical: it becomes coupled to a specific domain layer (closure equations / Richards-style formulation). The NestedNewtonThomas class also references the Casulli & Zanolli (2010) nested Newton approach, suggesting it may be more “process-specific” than a general-purpose solver.
Question: should we move NestedNewtonThomas to a more domain-oriented package/module (e.g., closure_equation), keeping numerical free of model-specific dependencies?
NestedNewtonThomas and NestedNewtonGC currently lives in it.geoframe.blogspot.numerical.newtonalgorithm, but it directly depends on it.geoframe.blogspot.closureequation.equationstate.EquationState.
Because of this dependency, the numerical module is no longer purely generic/numerical: it becomes coupled to a specific domain layer (closure equations / Richards-style formulation). The NestedNewtonThomas class also references the Casulli & Zanolli (2010) nested Newton approach, suggesting it may be more “process-specific” than a general-purpose solver.
Question: should we move NestedNewtonThomas to a more domain-oriented package/module (e.g., closure_equation), keeping numerical free of model-specific dependencies?