(eno Enhancement No.) (Place in code) (Suggested enhancement)
eno3 laopt::variable There should be a way to cast the value of a scalar variable (size 1x1) to double. So far, I can only write it to an Eigen::Matrix< > type, it seems.
eno5 eval_objective() It is not possible to pass a const Eigen object to this function for evaluation. I.e., it is not possible to directly pass a const reference to the function, e.g., a typical get function: const double obj_eval = opt_problem.eval_objective(laopt::Eval(), solver.sol_primal()).
eno6 laopt::variable Allow to use variable as scalar if length is one.
eno8 'casting to T' Simplify/make casting (of literals) to differentiable type T easier. | Allow explicit usage of math functions with namespace and/or template type: E.g., laopt::atan2( )
(eno Enhancement No.) (Place in code) (Suggested enhancement)
eno3
laopt::variableThere should be a way to cast the value of a scalar variable (size 1x1) to double. So far, I can only write it to an Eigen::Matrix< > type, it seems.eno5
eval_objective()It is not possible to pass a const Eigen object to this function for evaluation. I.e., it is not possible to directly pass a const reference to the function, e.g., a typical get function:const double obj_eval = opt_problem.eval_objective(laopt::Eval(), solver.sol_primal()).eno6
laopt::variableAllow to use variable as scalar if length is one.eno8 'casting to T' Simplify/make casting (of literals) to differentiable type T easier. | Allow explicit usage of math functions with namespace and/or template type: E.g., laopt::atan2( )