Fix non-linearity check during linear constraints

Fix non-linearity check during linear constraints to only evaluate variables with a non-zero influence in the constraint
Expand a test-case to also test that behavior

CADETProcess/optimization/optimizationProblem.py

@@ -37,7 +37,6 @@
 )
 
 from CADETProcess.metric import MetricBase
-
 from CADETProcess.optimization import Individual, Population
 from CADETProcess.optimization import ResultsCache
 
@@ -2170,7 +2169,7 @@ def A_transformed(self):
                 if isinstance(t, NoTransform):
                     continue
 
-                if not t.is_linear:
+                if a[j] != 0 and not t.is_linear:
                     raise CADETProcessError(
                         "Non-linear transform was used in linear constraints."
                     )
@@ -2250,7 +2249,7 @@ def b_transformed(self):
                 if isinstance(t, NoTransform):
                     continue
 
-                if not t.is_linear:
+                if a[j] != 0 and not t.is_linear:
                     raise CADETProcessError(
                         "Non-linear transform was used in linear constraints."
                     )
@@ -2447,7 +2446,7 @@ def Aeq_transformed(self):
                 if isinstance(t, NoTransform):
                     continue
 
-                if not t.is_linear:
+                if aeq[j] != 0 and not t.is_linear:
                     raise CADETProcessError(
                         "Non-linear transform was used in linear constraints."
                     )
@@ -2527,7 +2526,7 @@ def beq_transformed(self):
                 if isinstance(t, NoTransform):
                     continue
 
-                if not t.is_linear:
+                if aeq[j] != 0 and not t.is_linear:
                     raise CADETProcessError(
                         "Non-linear transform was used in linear constraints."
                     )

tests/optimization_problem_fixtures.py

@@ -165,29 +165,30 @@ def __init__(self, transform=None, has_evaluator=False, *args, **kwargs):
     def setup_variables(self, transform):
         self.add_variable('var_0', lb=-2, ub=2, transform=transform)
         self.add_variable('var_1', lb=-2, ub=2, transform=transform)
-
+        self.add_variable('var_2', lb=0, ub=2, transform="log")
     def setup_linear_constraints(self):
         self.add_linear_constraint(['var_0', 'var_1'], [-1, -0.5], 0)
 
     def _objective_function(self, x):
-        return x[0] - x[1]
+        return x[0] - x[1] + x[2]
 
     @property
     def optimal_solution(self):
-        x = np.array([-1, 2]).reshape(1, self.n_variables)
+        x = np.array([-1, 2, 0.0]).reshape(1, self.n_variables)
         f = -3
 
         return x, f
 
     @property
     def x0(self):
-        return [-0.5, 1.5]
+        return [-0.5, 1.5, 0.1]
 
     @property
     def conditional_minima(self):
         f_x0 = lambda x0:  x0 - 2
         f_x1 = lambda x1:  x1 * - 3/2
-        return f_x0, f_x1
+        f_x2 = lambda x2: x2
+        return f_x0, f_x1, f_x2
 
     def test_if_solved(self, optimization_results: OptimizationResults,
                        test_kwargs=default_test_kwargs):