small updates

marymorstan/cli.py

@@ -209,13 +209,13 @@ def get_cli_args():
         type=str
     )
 
-    parser.add_argument(
-        '--dataset-mapper-file',
-        action='store',
-        dest='dataset_mapper_file',
-        default=None,
-        type=str
-    )
+    # parser.add_argument(
+    #     '--dataset-mapper-file',
+    #     action='store',
+    #     dest='dataset_mapper_file',
+    #     default=None,
+    #     type=str
+    # )
 
     parser.add_argument(
         '--search-space',

marymorstan/evolutionary_algorithms/ea_space.py

@@ -32,7 +32,7 @@ class EASpace(Sequence):  # pylint: disable=too-many-public-methods
             try:
                 self.check_ea_space()
             except InvalidEASpace as e:
-                logging.error("EA space is not well formated: %s", e)
+                logging.error("EA space is not well formatted: %s", e)
                 raise e
 
     def set_parameters(self, parameters):
@@ -225,7 +225,8 @@ class EASpace(Sequence):  # pylint: disable=too-many-public-methods
             raise Exception(f"Unknown probability_operator_name {operator_type}")
 
         strategy = self._operator_strategy(operator_type)
-        probabilities = self.parameters[probability_operator_name] if probability_operator_name in self.parameters else None
+        probabilities = self.parameters[probability_operator_name] if (probability_operator_name
+                                                                       in self.parameters) else None
         len_operators = len(self._operators(operator_type))
 
         if strategy == OPERATOR_DISTRIBUTION_STRATEGY.STATIC_DISTRIBUTION:

marymorstan/marymorstan.py

@@ -103,7 +103,8 @@ class MaryMorstan():
         ea_space.set_parameters(ea_parameters)
 
         if (OPERATOR_DISTRIBUTION_STRATEGY.TIME_BASED_DISTRIBUTION in (ea_space.mutation_strategy,
-                                                                       ea_space.crossover_strategy) and wall_time_optimization_seconds is None):
+                                                                       ea_space.crossover_strategy)
+                and wall_time_optimization_seconds is None):
             raise Exception('Please specify wall_time_optimization_seconds parameters '
                             'if you plan to base your variation strategy on the time')
 

misc/integration-tests/regression.sh

@@ -15,6 +15,5 @@ then
   exit ${exit_code}
 fi
 
-
 echo ${results} | grep 0.11792351206144569 >/dev/null || exit 41
 echo ${results} | grep 0.7401839889682496 >/dev/null || exit 42

misc/simpleDemo.py

@@ -2,6 +2,7 @@
 
 from sklearn.model_selection import train_test_split
 from sklearn.model_selection._validation import _score
+import numpy as np
 
 import importlib
 import logging
@@ -11,27 +12,31 @@ from marymorstan.marymorstan import MaryMorstan
 log_level = getattr(logging, 'ERROR')
 logging.basicConfig(format='%(asctime)s [%(filename)s:%(lineno)d] %(levelname)s %(message)s', level=log_level)
 
-dataset_preprocessing_module = importlib.import_module("datasets.iris_dataset_preprocessing")
+dataset_preprocessing_module = importlib.import_module("datasets.iris_dataset_preprocessing")  # <1>
 dataset = dataset_preprocessing_module.MyDataSetPreprocessing("iris")
 
 X_train, X_test, y_train, y_test = train_test_split(dataset.get_X(), dataset.get_y(), test_size=.25, random_state=42)
 
-mm = MaryMorstan(generations=4, population_size=5, random_state=42)
-pipelines = mm.optimize(X_train=X_train, y_train=y_train, random_state=42)
+mm = MaryMorstan(generations=4, population_size=5, random_state=np.random.RandomState(42))  # <2>
+pipelines = mm.optimize(X_train=X_train, y_train=y_train,
+                        random_state=np.random.RandomState(42))  # <3>
 
-best_pipeline = MaryMorstan.best(pipelines)
+best_pipeline = MaryMorstan.best(pipelines)  # <4>
+# then you can save it as a string and easily reimport later
+best_pipeline_str = str(best_pipeline)
 
 print("best pipeline found:", str(best_pipeline))
 print("Objectives:", str(mm.objectives))
 
+print("resulting scores")  # <5>
+
 print(f'current validation_scores: {str(list(best_pipeline.fitness.weighted_values))}')
 
 best_pipeline_compiled = best_pipeline.compile()
 best_pipeline_compiled.fit(X_train, y_train)
 
 
-score = best_pipeline_compiled.score(X_train, y_train)
-print("train score 1: ", score)
+
 train_scores = _score(best_pipeline_compiled, X_train, y_train, mm.objectives.scorers).values()
 test_scores = _score(best_pipeline_compiled, X_test, y_test, mm.objectives.scorers).values()