more improvements to model opt + tests

Author: yoelcortes

biosteam/evaluation/_model.py

@@ -102,10 +102,12 @@ class Model:
     )
     default_optimizer_options = {
         'shgo': dict(f_tol=1e-3, minimizer_kwargs=dict(f_tol=1e-3)),
-        'differential evolution': {'seed': 0, 'popsize': 12, 'tol': 1e-3},
-        'COBYLA': {},
+        'differential evolution': dict(seed=0),
     }
-    default_optimizer = 'COBYLA'
+    for method in ('cobyla', 'cobyqa', 'trust-constr', 'slsqp', 'L-BFGS-B'):
+        default_optimizer_options[method] = {}
+        
+    default_optimizer = 'cobyla'
     default_convergence_model = None # Optional[str] Default convergence model
     load_default_parameters = load_default_parameters
 
@@ -365,7 +367,8 @@ def parameter(self,
         elif not setter:
             return lambda setter: self.parameter(setter, element, coupled, name,
                                                  distribution, units, baseline,
-                                                 bounds, hook, description, optimized)
+                                                 bounds, hook, description, optimized,
+                                                 kind, safe)
         p = Parameter(name, setter, element,
                       self.system, distribution, units, 
                       baseline, bounds, coupled, hook, description)
@@ -920,7 +923,7 @@ def optimize(self,
         for i, p in enumerate(parameters):
             lb[i], ub[i] = p.bounds
         bounds = Bounds(lb, ub)
-        if method == 'COBYLA':
+        if method in ('cobyla', 'cobyqa', 'trust-constr', 'slsqp', 'l-bfgs-b'):
             result = minimize(
                 objective_function,
                 args=args, 
@@ -939,7 +942,10 @@ def optimize(self,
             )
         else:
             raise ValueError(f'invalid optimization method {method!r}')
-        return result, convergence_model
+        if isinstance(convergence_model, str):
+            return result, convergence_model
+        else:
+            return result
 
     def evaluate(self, notify=0, file=None, autosave=0, autoload=False,
                  convergence_model=None, **kwargs):

biosteam/units/design_tools/Gibbs_equilibrium_reaction.py

@@ -81,7 +81,7 @@ def minimize_Gibbs_free_energy(
         polish = lambda *args, **kwargs: minimize(
             *args, **kwargs, 
             args=f_args,
-            method='COBYQA', 
+            method='COBYLA', 
         )
         solution = differential_evolution(
             Gibbs_equilibrium_objective,

tests/test_model.py

@@ -202,7 +202,7 @@ def set_temperature(temperature):
         H1.T = temperature
 
     with pytest.raises(ValueError):
-        @model.parameter(element=H1)
+        @model.parameter(element=H1, safe=True)
         def set_efficiency(temperature):
             H1.T = temperature
 
@@ -343,7 +343,7 @@ def test_model_optimization_differential_evolution():
     import biosteam as bst
     import numpy as np
     model = bst.Model(bst.System())
-    inputs = np.array([0, 0, 0])
+    inputs = np.array([0, 0], float)
 
     @model.optimized_parameter(bounds=(-2, 1))
     def P0(x0):
@@ -353,19 +353,20 @@ def P0(x0):
     def P1(x1):
         inputs[1] = x1
 
-    @model.optimized_parameter(bounds=(-1, 3))
-    def P2(x2):
-        inputs[2] = x2
-    
     @model.indicator
     def objective():
-        x0, x1, x2 = inputs
-        return x0**2 - x0 + x1**4 - x1**2 + x0 * x1 + x2
-    
-    solution = model.optimize(
-        objective, 
-        method='differential evolution',
-    )
+        x0, x1 = inputs
+        return x0**2 - x0 + x1**2 - x1
+    
+    for method in ('cobyla', 'cobyqa', 
+                   'trust-constr', 'slsqp',
+                   'L-BFGS-B', 'shgo',
+                   'differential evolution'):
+        solution = model.optimize(
+            objective, 
+            method=method,
+        )
+        assert_allclose(solution.x, [0.5, 0.5], rtol=1e-3, atol=1e-3)
 
 if __name__ == '__main__':
     test_parameter_hook()