It would be useful to support masks for selected variable and constraint instances after an ExaModel has been constructed.
My use case is defining a generic modeling template that is repeated across different units or sections of a chemical process flowsheet, while masking selected variables and constraints where parts of the template do not apply.
Conceptually, I am looking for something like:
model = ExaModel(core)
set_variable_mask!(model, x, variable_mask)
set_constraint_mask!(model, g, constraint_mask)
I am interested in masked evaluation rather than structural deletion, hence preserving the original model dimensions and sparsity pattern. Masked variables could be fixed at specified values, while masked constraints and their derivative contributions would ideally be skipped during evaluation.
It would be useful to support masks for selected variable and constraint instances after an ExaModel has been constructed.
My use case is defining a generic modeling template that is repeated across different units or sections of a chemical process flowsheet, while masking selected variables and constraints where parts of the template do not apply.
Conceptually, I am looking for something like:
I am interested in masked evaluation rather than structural deletion, hence preserving the original model dimensions and sparsity pattern. Masked variables could be fixed at specified values, while masked constraints and their derivative contributions would ideally be skipped during evaluation.