CyIpoptProblemInterface
(class from pyomo.contrib.pynumero.interfaces.cyipopt_interface)
- class pyomo.contrib.pynumero.interfaces.cyipopt_interface.CyIpoptProblemInterface[source]
Bases:
objectAbstract subclass of
cyipopt.Problemdefining an object that can be used as an interface to CyIpopt. Subclasses must define all methods necessary for the CyIpopt solve and must call this class’s__init__method to initialize Ipopt’s data structures.Note that, if “output_file” is provided as an Ipopt option, the log file is open until this object (and thus the underlying Ipopt NLP object) is deallocated. To force this deallocation, call the
close()method, which is defined bycyipopt.Problem.- __init__()[source]
Initialize the problem interface
This method calls
cyipopt.Problem.__init__, and must be called by any subclass’s__init__method. If not, we will segfault when we callcyipopt.Problem.solvefrom this object.
Methods
__init__()Initialize the problem interface
constraints(x)Return the residuals of the constraints evaluated at x as a numpy ndarray
g_lb()Return the lower bounds on the constraints as a numpy ndarray
g_ub()Return the upper bounds on the constraints as a numpy ndarray
gradient(x)Return the gradient of the objective function evaluated at x as a numpy ndarray
hessian(x, y, obj_factor)Return the values for the hessian evaluated at x as a numpy ndarray of nonzero values corresponding to the rows and columns specified in the hessianstructure method.
Return the structure of the hessian in coordinate format.
intermediate(alg_mod, iter_count, obj_value, ...)Callback that can be used to examine or report intermediate results.
jacobian(x)Return the values for the jacobian evaluated at x as a numpy ndarray of nonzero values corresponding to the rows and columns specified in the jacobianstructure
Return the structure of the jacobian in coordinate format.
objective(x)Return the value of the objective function evaluated at x
Return the values for scaling factors as a tuple (objective_scaling, x_scaling, g_scaling).
solve(x[, lagrange, zl, zu])Solve a CyIpopt Problem
x_init()Return the initial values for x as a numpy ndarray
x_lb()Return the lower bounds on x as a numpy ndarray
x_ub()Return the upper bounds on x as a numpy ndarray
Member Documentation
- abstract constraints(x)[source]
Return the residuals of the constraints evaluated at x as a numpy ndarray
- abstract gradient(x)[source]
Return the gradient of the objective function evaluated at x as a numpy ndarray
- abstract hessian(x, y, obj_factor)[source]
Return the values for the hessian evaluated at x as a numpy ndarray of nonzero values corresponding to the rows and columns specified in the hessianstructure method. Note: return ONLY the lower diagonal of this symmetric matrix.
- abstract hessianstructure()[source]
Return the structure of the hessian in coordinate format. That is, return (rows,cols) where rows and cols are both numpy ndarray objects that contain the row and column indices for each of the nonzeros in the hessian. Note: return ONLY the lower diagonal of this symmetric matrix.
- intermediate(alg_mod, iter_count, obj_value, inf_pr, inf_du, mu, d_norm, regularization_size, alpha_du, alpha_pr, ls_trials)[source]
Callback that can be used to examine or report intermediate results. This method is called each iteration
- abstract jacobian(x)[source]
Return the values for the jacobian evaluated at x as a numpy ndarray of nonzero values corresponding to the rows and columns specified in the jacobianstructure
- abstract jacobianstructure()[source]
Return the structure of the jacobian in coordinate format. That is, return (rows,cols) where rows and cols are both numpy ndarray objects that contain the row and column indices for each of the nonzeros in the jacobian.
- abstract scaling_factors()[source]
Return the values for scaling factors as a tuple (objective_scaling, x_scaling, g_scaling). Return None if the scaling factors are to be ignored