refactor base-classes for test-functions

src/surfaces/test_functions/_base_test_function.py

@@ -4,6 +4,11 @@
 
 import time
 
+import numpy as np
+import pandas as pd
+
+from ..data_collector import SurfacesDataCollector
+
 
 class BaseTestFunction:
     explanation = """ """
@@ -15,9 +20,20 @@ class BaseTestFunction:
     objective_function: callable
     pure_objective_function: callable
 
-    def __init__(self):
+    def __init__(self, metric, sleep=0, evaluate_from_data=False):
+        self.sleep = sleep
+        self.metric = metric
+
         self.create_objective_function()
 
+        if evaluate_from_data:
+            self.sdc = SurfacesDataCollector()
+            self.objective_function = self.objective_function_loaded
+        else:
+            self.objective_function = self.pure_objective_function
+
+        self.objective_function.__name__ = self.__name__
+
     def create_objective_function(self):
         e_msg = "'create_objective_function'-method is not implemented"
         raise NotImplementedError(e_msg)
@@ -26,9 +42,6 @@ def search_space(self):
         e_msg = "'search_space'-method is not implemented"
         raise NotImplementedError(e_msg)
 
-    def load_search_data(self):
-        return self.sql_data.load(self.__name__)
-
     def return_metric(self, loss):
         if self.metric == "score":
             return -loss
@@ -52,3 +65,28 @@ def objective_function(self, *input):
             metric = self.objective_function_np(*input)
 
         return self.return_metric(metric)
+
+    def objective_function_loaded(self, params):
+        try:
+            parameter_d = params.para_dict
+        except AttributeError:
+            parameter_d = params
+
+        for para_names, dim_value in parameter_d.items():
+            try:
+                parameter_d[para_names] = dim_value.__name__
+            except AttributeError:
+                pass
+
+        search_data = self.sdc.load(self.__name__)
+        if search_data is None:
+            msg = "Search Data is empty"
+            raise TypeError(msg)
+
+        params_df = pd.DataFrame(parameter_d, index=[0])
+
+        para_df_row = search_data[
+            np.all(search_data[self.para_names].values == params_df.values, axis=1)
+        ]
+        score = para_df_row["score"].values[0]
+        return score

src/surfaces/test_functions/machine_learning/base_machine_learning_function.py

@@ -3,52 +3,9 @@
 # License: MIT License
 
 
-import numpy as np
-import pandas as pd
-from functools import reduce
-
-from hyperactive import Hyperactive
-from hyperactive.optimizers import GridSearchOptimizer
 from .._base_test_function import BaseTestFunction
 
 
 class MachineLearningFunction(BaseTestFunction):
-    def __init__(self, metric=None, sleep=0, load_search_data=False):
-        super().__init__()
-
-        self.metric = metric
-        self.sleep = sleep
-
-        self.create_objective_function()
-
-        if load_search_data:
-            self.objective_function = self.objective_function_loaded
-        else:
-            self.objective_function = self.pure_objective_function
-
-        # self.objective_function.__func__.__name__ = self.__name__
-
-    def objective_function_loaded(self, params):
-        try:
-            parameter_d = params.para_dict
-        except AttributeError:
-            parameter_d = params
-
-        for para_names, dim_value in parameter_d.items():
-            try:
-                parameter_d[para_names] = dim_value.__name__
-            except AttributeError:
-                pass
-
-        search_data = self.load_search_data()
-        if search_data is None:
-            msg = "Search Data is empty"
-            raise TypeError(msg)
-
-        params_df = pd.DataFrame(parameter_d, index=[0])
-
-        para_df_row = search_data[
-            np.all(search_data[self.para_names].values == params_df.values, axis=1)
-        ]
-        score = para_df_row["score"].values[0]
-        return score
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)

src/surfaces/test_functions/mathematical/_base_objective_function.py

@@ -4,10 +4,6 @@
 
 
 import numpy as np
-import pandas as pd
-from functools import reduce
-
-from gradient_free_optimizers import GridSearchOptimizer
 
 from .._base_test_function import BaseTestFunction
 
@@ -19,8 +15,15 @@ class MathematicalFunction(BaseTestFunction):
     formula = r" "
     global_minimum = r" "
 
-    def __init__(self, metric="score", input_type="dictionary", sleep=0):
-        super().__init__()
+    def __init__(
+        self,
+        *args,
+        metric="loss",
+        input_type="dictionary",
+        sleep=0,
+        **kwargs,
+    ):
+        super().__init__(*args, metric, **kwargs)
 
         self.metric = metric
         self.input_type = input_type