Re-remove accidentally readded exit conditions

process/core/solver/solver.py

@@ -325,25 +325,13 @@ def verror(self):
                     "A feasible solution may be difficult to achieve.",
                     "Try changing or adding variables to IXC.",
                 ),
-                4: (
-                    "An uphill search direction was found.",
-                    "Try changing the equations in ICC, or",
-                    "adding new variables to IXC.",
-                ),
                 SolverOutputCondition.NO_SOLUTION: (
                     "The quadratic programming technique was unable to",
                     "find a feasible point.\n",
                     "Try changing or adding variables to IXC, or modify",
                     "their initial values (especially if only 1 optimisation",
                     "iteration was performed).",
                 ),
-                6: (
-                    "The quadratic programming technique was restricted",
-                    "by an artificial bound, or failed due to a singular",
-                    "matrix.",
-                    "Try changing the equations in ICC, or",
-                    "adding new variables to IXC.",
-                ),
             }.get(self.info, "Unknown Error code")
         )
 

process/core/solver/solver_handler.py

@@ -1,6 +1,7 @@
 """Module containing solver handler routines"""
 
 import logging
+from contextlib import contextmanager
 
 from tabulate import tabulate
 
@@ -50,14 +51,9 @@ def run(self):
         # Initialise iteration variables and bounds in Python: relies on Fortran
         # iteration variables being defined above
         # Trim maximum size arrays down to actually used size
-        n = self.data.numerics.nvar
-        x = self.data.numerics.xcm[:n]
-        bndl = self.data.numerics.itv_scaled_lower_bounds[:n]
-        bndu = self.data.numerics.itv_scaled_upper_bounds[:n]
-
-        # Define total number of constraints and equality constraints
-        m = self.data.numerics.neqns + self.data.numerics.nineqns
-        meq = self.data.numerics.neqns
+        x = self.data.numerics.xcm[: self.data.numerics.nvar]
+        bndl = self.data.numerics.itv_scaled_lower_bounds[: self.data.numerics.nvar]
+        bndu = self.data.numerics.itv_scaled_upper_bounds[: self.data.numerics.nvar]
 
         # Evaluators() calculates the objective and constraint functions and
         # their gradients for a given vector x
@@ -68,31 +64,21 @@ def run(self):
         self.solver.set_evaluators(evaluators)
         self.solver.set_bounds(bndl, bndu)
         self.solver.set_opt_params(x)
-        self.solver.set_constraints(m, meq)
+        # Define total number of constraints and equality constraints
+        self.solver.set_constraints(
+            m=self.data.numerics.neqns + self.data.numerics.nineqns,
+            meq=self.data.numerics.neqns,
+        )
         ifail = self.solver.solve()
 
         # If VMCON optimisation has failed then try altering value of epsfcn
         if self.solver_name == "vmcon":
             if ifail != SolverOutputCondition.CONVERGED:
-                print("Trying again with new epsfcn")
-                # epsfcn is only used in evaluators.Evaluators()
-                # TODO epsfcn could be set in Evaluators instance now, don't need to
-                # set/unset in self.data.numerics module
-                self.data.numerics.epsfcn *= 10  # try new larger value
-                print("new epsfcn = ", self.data.numerics.epsfcn)
-
-                ifail = self.solver.solve()
-                # First solution attempt failed
-                # (ifail != SolverOutputCondition.CONVERGED): supply ifail value
-                # to next attempt
-                self.data.numerics.epsfcn /= 10  # reset value
-
+                with epsfcn_context(self.data.numerics, 10):
+                    ifail = self.solver.solve()
             if ifail != SolverOutputCondition.CONVERGED:
-                print("Trying again with new epsfcn")
-                self.data.numerics.epsfcn /= 10  # try new smaller value
-                print("new epsfcn = ", self.data.numerics.epsfcn)
-                ifail = self.solver.solve()
-                self.data.numerics.epsfcn *= 10  # reset value
+                with epsfcn_context(self.data.numerics, 0.1):
+                    ifail = self.solver.solve()
 
             # If VMCON has exited with error code 5
             # (ifail = SolverOutputCondition.NO_SOLUTION) try another run using a
@@ -346,3 +332,18 @@ def _optimisation_parameters_output(self):
                 numalign="left",
             ),
         )
+
+
+@contextmanager
+def epsfcn_context(numerics, factor):
+    """Set and then reset epsfcn value"""
+    print("Trying again with new epsfcn")
+    # epsfcn is only used in evaluators.Evaluators()
+    # TODO epsfcn could be set in Evaluators instance now, don't need to
+    # set/unset in numerics module
+    numerics.epsfcn *= factor  # try new larger value
+    print("new epsfcn = ", numerics.epsfcn)
+    try:
+        yield
+    finally:
+        numerics.epsfcn /= factor  # reset value