Revert "FWPH: Handling LP warmstart better"

Author: bknueven

mpisppy/opt/fwph.py

@@ -150,22 +150,17 @@ def fwph_main(self, finalize=True):
                 tee=teeme,
                 verbose=self.options["verbose"],
                 # sdm_iter_limit=20,
-                # don't cut off integer solutions
-                # for this pass
-                FW_conv_thresh=1e+10,
+                # FW_conv_thresh=1e-8,
             )
+            global_toc("Starting FW PH")
 
-        # sometimes we take a while to initialize,
-        # sometimes LP may prove optimizality
-        # check before entering the main loop
-        if self.spcomm and self.spcomm.is_converged():
-            if finalize:
-                weight_dict = self._gather_weight_dict() # None if rank != 0
-                xbars_dict  = self._get_xbars() # None if rank != 0
-                return 0, weight_dict, xbars_dict
-            return 0
-
-        global_toc("Starting FW PH")
+        else:
+            # FWPH can take some time to initialize
+            # If run as a spoke, check for convergence here
+            if self.spcomm and self.spcomm.is_converged():
+                if finalize:
+                    return 0, None, None
+                return 0
 
         self.iterk_loop()
 
@@ -406,7 +401,7 @@ def SDM(self, model_name, mip_solver_options, dtiming, tee, verbose, sdm_iter_li
                         -  scen_mip._mpisppy_model.xbars[ndn_i]._value))
 
             self._fix_fixings(model_name, mip, qp)
-            cutoff = self._add_objective_cutoff(mip, qp, model_name, best_bound_update, FW_conv_thresh)
+            cutoff = self._add_objective_cutoff(mip, qp, model_name, best_bound_update)
             # print(f"{model_name=}, {cutoff=}")
             # Algorithm 2 line 5
             self.solve_one(
@@ -586,15 +581,15 @@ def _compute_gamma_t(self, cutoff, inner_bound):
             print('Try decreasing the MIP gap tolerance and re-solving')
         return stop_check
 
-    def _add_objective_cutoff(self, mip, qp, model_name, best_bound_update, FW_conv_thresh):
+    def _add_objective_cutoff(self, mip, qp, model_name, best_bound_update):
         """ Add a constraint to the MIP objective ensuring
             an improving direction in the QP subproblem is generated
         """
         assert not hasattr(mip._mpisppy_model, "obj_cutoff_constraint")
         # print(f"\tnonants part: {pyo.value(qp._mpisppy_model.mip_obj_in_qp)}")
         # print(f"\trecoursepart: {pyo.value(qp.recourse_cost)}")
         cutoff = pyo.value(qp._mpisppy_model.mip_obj_in_qp) + pyo.value(qp.recourse_cost)
-        epsilon = FW_conv_thresh
+        epsilon = self.FW_options.get("stop_check_tol", 1e-4)
         # normalized Gamma^t
         epsilon = max(epsilon, abs(epsilon*cutoff))
         # tbmipsolve = time.perf_counter()