PEP8 (1)

rl_core/experiments/ts_experiment_dqn.py

@@ -27,7 +27,11 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
     )
 
     dataloader_train, dataloader_test, train_list, test_list = define_data_for_experiment()
-    env = TimeSeriesPipelineEnvironment(max_number_of_nodes=number_of_nodes_in_pipeline, render_mode='none', metadata_dim=126)
+    env = TimeSeriesPipelineEnvironment(
+        max_number_of_nodes=number_of_nodes_in_pipeline,
+        render_mode='none',
+        metadata_dim=126
+    )
     state_dim, action_dim = env.state_dim, env.action_dim
 
     agent = DQN(

rl_core/experiments/ts_validation_dqn.py

@@ -13,7 +13,8 @@
 
 
 def print_params(experiment_name, number_of_nodes_in_pipeline=8):
-    log_dir = f'{project_root()}/MetaFEDOT/rl_core/agent/tensorboard_logs/dqn/{number_of_nodes_in_pipeline}/{experiment_name}'
+    log_dir = f'{project_root()}/MetaFEDOT/rl_core/agent/tensorboard_logs/dqn/' \
+              f'{number_of_nodes_in_pipeline}/{experiment_name}'
 
     with io.open(f'{log_dir}/params.log', 'r', encoding='utf-8') as file:
         lines = file.readlines()
@@ -57,8 +58,6 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
     total_metrics = []
 
     period = 20
-    period_of_cleaning = 15
-    period_of_heatmap = 100
 
     for episode in range(1, n_episodes + 1):
         print(f'-- Starting {episode} episode --')
@@ -119,15 +118,16 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
         print(f'-- Finishing {episode} episode --\n')
 
     # -- Saving Agent ---
-    name = f'{env.metadata["name"]}_{number_of_nodes_in_pipeline}_{state_dim}_{agent.metadata["name"]}_{agent.hidden_dim}_{n_episodes}'
+    name = f'{env.metadata["name"]}_{number_of_nodes_in_pipeline}_{state_dim}_{agent.metadata["name"]}' \
+           f'_{agent.hidden_dim}_{n_episodes}'
     path = f'{log_dir}/weight'
 
     if not os.path.exists(path):
         os.makedirs(path)
 
     agent.save(f'{path}/{name}')
 
-    print(f'-- Validation Starts --\n')
+    print('-- Validation Starts --\n')
     for test_dataset in test_list:
         train_data, test_data, meta_data = dataloader_test.get_data(test_dataset)
 
@@ -192,6 +192,3 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
 
     for h, g, e, e_m, e_d in params:
         run_experiment(2000, 8, h, g, e, e_m, e_d)
-
-    # Parallel(n_jobs=-2)(
-    #     delayed(run_experiment)(2000, 8, h, g, e, e_m, e_d) for m, h, g, e, e_m, e_d in params)

rl_core/experiments/ts_validation_dt.py

@@ -127,7 +127,7 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
 
     agent.save(f'{path}/{name}')
 
-    print(f'-- Validation Starts --\n')
+    print('-- Validation Starts --\n')
     for test_dataset in test_list:
         train_data, test_data, meta_data = dataloader_test.get_data(test_dataset)
 
@@ -191,4 +191,4 @@ def run_experiment(n_episodes, number_of_nodes_in_pipeline, hidden_dim, gamma, e
     ]
 
     for h, g, e, e_m, e_d in params:
-        run_experiment(2000, 8, h, g, e, e_m, e_d)
+        run_experiment(2000, 8, h, g, e, e_m, e_d)

rl_core/pipeline_validation.py

@@ -102,7 +102,8 @@
 
         path_to_base = os.path.join(
             str(project_root()),
-            f'MetaFEDOT\\data\\knowledge_base_time_series_0\\datasets\\{dataset}\\model\\0_pipeline_saved\\0_pipeline_saved.json'
+            f'MetaFEDOT\\data\\knowledge_base_time_series_0\\datasets\\{dataset}\\'
+            f'model\\0_pipeline_saved\\0_pipeline_saved.json'
         )
 
         try:
@@ -121,7 +122,8 @@
         try:
             path_to_topo = os.path.join(
                 str(project_root()),
-                f'MetaFEDOT\\data\\topo_ws_selection_evo\\{dataset.split("_")[1]}\\model\\0_pipeline_saved\\0_pipeline_saved.json'
+                f'MetaFEDOT\\data\\topo_ws_selection_evo\\{dataset.split("_")[1]}\\'
+                f'model\\0_pipeline_saved\\0_pipeline_saved.json'
             )
 
             topo_pipeline_ = Pipeline().load(source=path_to_topo)