add support for cyipopt

doc/OnlineDocs/contributed_packages/mindtpy.rst

@@ -18,6 +18,10 @@ The following algorithms are currently available in MindtPy:
 Usage and early implementation details for MindtPy can be found in the PSE 2018 paper Bernal et al.,
 (`ref <https://doi.org/10.1016/B978-0-444-64241-7.50144-0>`_,
 `preprint <http://egon.cheme.cmu.edu/Papers/Bernal_Chen_MindtPy_PSE2018Paper.pdf>`_).
+This solver implementation has been developed by `David Bernal <https://github.com/bernalde>`_
+and `Zedong Peng <https://github.com/ZedongPeng>`_ as part of research efforts at the `Bernal Research Group 
+<https://bernalde.github.io/>`_ and the `Grossmann Research Group <http://egon.cheme.cmu.edu/>`_
+at Purdue University and Carnegie Mellon University.
 
 .. _Duran & Grossmann, 1986: https://dx.doi.org/10.1007/BF02592064
 .. _Westerlund & Petterson, 1995: http://dx.doi.org/10.1016/0098-1354(95)87027-X
@@ -310,6 +314,6 @@ Report a Bug
 If you find a bug in MindtPy, we will be grateful if you could
 
 - submit an `issue`_ in Pyomo repository  
-- directly contact David Bernal <bernalde@cmu.edu> and Zedong Peng <peng_zedong@126.com>.
+- directly contact David Bernal <dbernaln@purdue.edu> and Zedong Peng <zdpeng95@gmail.com>.
 
 .. _issue: https://github.com/Pyomo/pyomo/issues

pyomo/contrib/mindtpy/MindtPy.py

@@ -78,11 +78,6 @@ class MindtPySolver(object):
     - Global LP/NLP based branch-and-bound (GLP/NLP)
     - Regularized LP/NLP based branch-and-bound (RLP/NLP)
     - Feasibility pump (FP)
-
-    This solver implementation has been developed by David Bernal <https://github.com/bernalde>
-    and Zedong Peng <https://github.com/ZedongPeng> as part of research efforts at the Grossmann
-    Research Group (http://egon.cheme.cmu.edu/) at the Department of Chemical Engineering at
-    Carnegie Mellon University.
     """
 
     CONFIG = _get_MindtPy_config()

pyomo/contrib/mindtpy/algorithm_base_class.py

@@ -806,7 +806,7 @@ def set_up_solve_data(self, model, config):
         self.primal_bound_progress = [self.primal_bound]
         self.dual_bound_progress = [self.dual_bound]
 
-        if config.nlp_solver == 'ipopt':
+        if config.nlp_solver in {'ipopt', 'cyipopt'}:
             if not hasattr(self.working_model, 'ipopt_zL_out'):
                 self.working_model.ipopt_zL_out = Suffix(direction=Suffix.IMPORT)
             if not hasattr(self.working_model, 'ipopt_zU_out'):
@@ -900,11 +900,22 @@ def init_rNLP(self, config, add_oa_cuts=True):
             )
             # Add OA cut
             if add_oa_cuts:
-                dual_values = (
-                    list(m.dual[c] for c in MindtPy.constraint_list)
-                    if config.calculate_dual_at_solution
-                    else None
-                )
+                if (
+                    self.config.nlp_solver == 'cyipopt'
+                    and self.objective_sense == minimize
+                ):
+                    # TODO: recover the opposite dual when cyipopt issue #2831 is solved.
+                    dual_values = (
+                        list(-1 * m.dual[c] for c in MindtPy.constraint_list)
+                        if config.calculate_dual_at_solution
+                        else None
+                    )
+                else:
+                    dual_values = (
+                        list(m.dual[c] for c in MindtPy.constraint_list)
+                        if config.calculate_dual_at_solution
+                        else None
+                    )
                 copy_var_list_values(
                     m.MindtPy_utils.variable_list,
                     self.mip.MindtPy_utils.variable_list,
@@ -1194,6 +1205,12 @@ def handle_subproblem_optimal(self, fixed_nlp, config, cb_opt=None, fp=False):
             for c in fixed_nlp.tmp_duals:
                 if fixed_nlp.dual.get(c, None) is None:
                     fixed_nlp.dual[c] = fixed_nlp.tmp_duals[c]
+                elif (
+                    self.config.nlp_solver == 'cyipopt'
+                    and self.objective_sense == minimize
+                ):
+                    # TODO: recover the opposite dual when cyipopt issue #2831 is solved.
+                    fixed_nlp.dual[c] = -fixed_nlp.dual[c]
             dual_values = list(
                 fixed_nlp.dual[c] for c in fixed_nlp.MindtPy_utils.constraint_list
             )
@@ -1798,7 +1815,7 @@ def set_up_mip_solver(self, config, regularization_problem=False):
             The customized MIP solver.
         """
         # Deactivate extraneous IMPORT/EXPORT suffixes
-        if config.nlp_solver == 'ipopt':
+        if config.nlp_solver in {'ipopt', 'cyipopt'}:
             getattr(self.mip, 'ipopt_zL_out', _DoNothing()).deactivate()
             getattr(self.mip, 'ipopt_zU_out', _DoNothing()).deactivate()
         if regularization_problem:

pyomo/contrib/mindtpy/config_options.py

@@ -508,7 +508,7 @@ def _add_subsolver_configs(CONFIG):
         'nlp_solver',
         ConfigValue(
             default='ipopt',
-            domain=In(['ipopt', 'appsi_ipopt', 'gams', 'baron']),
+            domain=In(['ipopt', 'appsi_ipopt', 'gams', 'baron', 'cyipopt']),
             description='NLP subsolver name',
             doc='Which NLP subsolver is going to be used for solving the nonlinear'
             'subproblems.',

pyomo/contrib/mindtpy/extended_cutting_plane.py

@@ -24,27 +24,16 @@
 @SolverFactory.register(
     'mindtpy.ecp', doc='MindtPy: Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo'
 )
-class MindtPy_OA_Solver(_MindtPyAlgorithm):
+class MindtPy_ECP_Solver(_MindtPyAlgorithm):
     """
     Decomposition solver for Mixed-Integer Nonlinear Programming (MINLP) problems.
 
     The MindtPy (Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo) solver
     applies a variety of decomposition-based approaches to solve Mixed-Integer
     Nonlinear Programming (MINLP) problems.
-    These approaches include:
-
-    - Outer approximation (OA)
-    - Global outer approximation (GOA)
-    - Regularized outer approximation (ROA)
-    - LP/NLP based branch-and-bound (LP/NLP)
-    - Global LP/NLP based branch-and-bound (GLP/NLP)
-    - Regularized LP/NLP based branch-and-bound (RLP/NLP)
-    - Feasibility pump (FP)
-
-    This solver implementation has been developed by David Bernal <https://github.com/bernalde>
-    and Zedong Peng <https://github.com/ZedongPeng> as part of research efforts at the Grossmann
-    Research Group (http://egon.cheme.cmu.edu/) at the Department of Chemical Engineering at
-    Carnegie Mellon University.
+    This class includes:
+
+    - Extended Cutting Plane (ECP)
     """
 
     CONFIG = _get_MindtPy_ECP_config()

pyomo/contrib/mindtpy/feasibility_pump.py

@@ -37,20 +37,9 @@ class MindtPy_FP_Solver(_MindtPyAlgorithm):
     The MindtPy (Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo) solver
     applies a variety of decomposition-based approaches to solve Mixed-Integer
     Nonlinear Programming (MINLP) problems.
-    These approaches include:
+    This class includes:
 
-    - Outer approximation (OA)
-    - Global outer approximation (GOA)
-    - Regularized outer approximation (ROA)
-    - LP/NLP based branch-and-bound (LP/NLP)
-    - Global LP/NLP based branch-and-bound (GLP/NLP)
-    - Regularized LP/NLP based branch-and-bound (RLP/NLP)
     - Feasibility pump (FP)
-
-    This solver implementation has been developed by David Bernal <https://github.com/bernalde>
-    and Zedong Peng <https://github.com/ZedongPeng> as part of research efforts at the Grossmann
-    Research Group (http://egon.cheme.cmu.edu/) at the Department of Chemical Engineering at
-    Carnegie Mellon University.
     """
 
     CONFIG = _get_MindtPy_FP_config()

pyomo/contrib/mindtpy/global_outer_approximation.py

@@ -45,27 +45,17 @@
 @SolverFactory.register(
     'mindtpy.goa', doc='MindtPy: Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo'
 )
-class MindtPy_OA_Solver(_MindtPyAlgorithm):
+class MindtPy_GOA_Solver(_MindtPyAlgorithm):
     """
     Decomposition solver for Mixed-Integer Nonlinear Programming (MINLP) problems.
 
     The MindtPy (Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo) solver
     applies a variety of decomposition-based approaches to solve Mixed-Integer
     Nonlinear Programming (MINLP) problems.
-    These approaches include:
+    This class includes:
 
-    - Outer approximation (OA)
     - Global outer approximation (GOA)
-    - Regularized outer approximation (ROA)
-    - LP/NLP based branch-and-bound (LP/NLP)
     - Global LP/NLP based branch-and-bound (GLP/NLP)
-    - Regularized LP/NLP based branch-and-bound (RLP/NLP)
-    - Feasibility pump (FP)
-
-    This solver implementation has been developed by David Bernal <https://github.com/bernalde>
-    and Zedong Peng <https://github.com/ZedongPeng> as part of research efforts at the Grossmann
-    Research Group (http://egon.cheme.cmu.edu/) at the Department of Chemical Engineering at
-    Carnegie Mellon University.
     """
 
     CONFIG = _get_MindtPy_GOA_config()

pyomo/contrib/mindtpy/mip_solve.py

@@ -176,7 +176,7 @@ def set_up_mip_solver(solve_data, config, regularization_problem):
         The customized MIP solver.
     """
     # Deactivate extraneous IMPORT/EXPORT suffixes
-    if config.nlp_solver == 'ipopt':
+    if config.nlp_solver in {'ipopt', 'cyipopt'}:
         getattr(solve_data.mip, 'ipopt_zL_out', _DoNothing()).deactivate()
         getattr(solve_data.mip, 'ipopt_zU_out', _DoNothing()).deactivate()
     if regularization_problem:

pyomo/contrib/mindtpy/nlp_solve.py

@@ -185,6 +185,12 @@ def handle_subproblem_optimal(fixed_nlp, solve_data, config, cb_opt=None):
         for c in fixed_nlp.tmp_duals:
             if fixed_nlp.dual.get(c, None) is None:
                 fixed_nlp.dual[c] = fixed_nlp.tmp_duals[c]
+            elif (
+                config.nlp_solver == 'cyipopt'
+                and solve_data.objective_sense == minimize
+            ):
+                # TODO: recover the opposite dual when cyipopt issue #2831 is solved.
+                fixed_nlp.dual[c] = -fixed_nlp.dual[c]
         dual_values = list(
             fixed_nlp.dual[c] for c in fixed_nlp.MindtPy_utils.constraint_list
         )

pyomo/contrib/mindtpy/outer_approximation.py

@@ -29,20 +29,12 @@ class MindtPy_OA_Solver(_MindtPyAlgorithm):
     The MindtPy (Mixed-Integer Nonlinear Decomposition Toolbox in Pyomo) solver
     applies a variety of decomposition-based approaches to solve Mixed-Integer
     Nonlinear Programming (MINLP) problems.
-    These approaches include:
+    This class includes:
 
     - Outer approximation (OA)
-    - Global outer approximation (GOA)
     - Regularized outer approximation (ROA)
     - LP/NLP based branch-and-bound (LP/NLP)
-    - Global LP/NLP based branch-and-bound (GLP/NLP)
     - Regularized LP/NLP based branch-and-bound (RLP/NLP)
-    - Feasibility pump (FP)
-
-    This solver implementation has been developed by David Bernal <https://github.com/bernalde>
-    and Zedong Peng <https://github.com/ZedongPeng> as part of research efforts at the Grossmann
-    Research Group (http://egon.cheme.cmu.edu/) at the Department of Chemical Engineering at
-    Carnegie Mellon University.
     """
 
     CONFIG = _get_MindtPy_OA_config()

pyomo/contrib/mindtpy/single_tree.py

@@ -498,6 +498,12 @@ def handle_lazy_subproblem_optimal(self, fixed_nlp, solve_data, config, opt):
             for c in fixed_nlp.tmp_duals:
                 if fixed_nlp.dual.get(c, None) is None:
                     fixed_nlp.dual[c] = fixed_nlp.tmp_duals[c]
+                elif (
+                    config.nlp_solver == 'cyipopt'
+                    and solve_data.objective_sense == minimize
+                ):
+                    # TODO: recover the opposite dual when cyipopt issue #2831 is solved.
+                    fixed_nlp.dual[c] = -fixed_nlp.dual[c]
             dual_values = list(
                 fixed_nlp.dual[c] for c in fixed_nlp.MindtPy_utils.constraint_list
             )

pyomo/contrib/mindtpy/util.py

@@ -565,6 +565,9 @@ def set_solver_options(opt, timing, config, solver_type, regularization=False):
     elif solver_name in {'ipopt', 'appsi_ipopt'}:
         opt.options['max_cpu_time'] = remaining
         opt.options['constr_viol_tol'] = config.zero_tolerance
+    elif solver_name == 'cyipopt':
+        opt.config.options['max_cpu_time'] = float(remaining)
+        opt.config.options['constr_viol_tol'] = config.zero_tolerance
     elif solver_name == 'gams':
         if solver_type == 'mip':
             opt.options['add_options'] = [