03101503

models_processing/model_koi/bp.yaml

@@ -2,7 +2,7 @@ Model:
   add_train: false
   batch_size: 64
   dropout_rate: 0.2
-  epoch: 100
+  epoch: 200
   fusion: true
   hidden_size: 64
   his_points: 16

models_processing/model_koi/cnn.yaml

@@ -2,7 +2,7 @@ Model:
   add_train: false
   batch_size: 64
   dropout_rate: 0.2
-  epoch: 100
+  epoch: 200
   fusion: true
   hidden_size: 64
   his_points: 16

models_processing/model_koi/lstm.yaml

@@ -2,7 +2,7 @@ Model:
   add_train: false
   batch_size: 64
   dropout_rate: 0.2
-  epoch: 100
+  epoch: 200
   fusion: true
   hidden_size: 64
   his_points: 16

models_processing/model_koi/tf_bp_pre.py

@@ -59,11 +59,13 @@ def model_prediction_bp():
         res = list(chain.from_iterable(target_scaler.inverse_transform([bp.predict(scaled_pre_x).flatten()])))
         pre_data['power_forecast'] = res[:len(pre_data)]
         pre_data['farm_id'] = 'J00083'
-        pre_data['cdq'] = 1
-        pre_data['dq'] = 1
-        pre_data['zq'] = 1
+        pre_data['cdq'] = args.get('cdq', 1)
+        pre_data['dq'] = args.get('dq', 1)
+        pre_data['zq'] = args.get('zq', 1)
+        res_cols = ['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']
+        res_cols += [args['target']] if args['algorithm_test'] else res_cols
         pre_data.rename(columns={args['col_time']: 'date_time'}, inplace=True)
-        pre_data = pre_data[['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']]
+        pre_data = pre_data[res_cols]
 
         pre_data['power_forecast'] = pre_data['power_forecast'].round(2)
         pre_data.loc[pre_data['power_forecast'] > g.opt.cap, 'power_forecast'] = g.opt.cap

models_processing/model_koi/tf_bp_train.py

@@ -18,7 +18,6 @@ import matplotlib.pyplot as plt
 from common.logs import Log
 logger = Log('tf_bp').logger
 np.random.seed(42)  # NumPy随机种子
-# tf.set_random_seed(42)  # TensorFlow随机种子
 app = Flask('tf_bp_train——service')
 
 with app.app_context():
@@ -26,7 +25,6 @@ with app.app_context():
         args = yaml.safe_load(f)
     dh = DataHandler(logger, args)
     bp = BPHandler(logger, args)
-    global opt
 
 @app.before_request
 def update_config():

models_processing/model_koi/tf_cnn_pre.py

@@ -60,11 +60,13 @@ def model_prediction_bp():
         res = list(chain.from_iterable(target_scaler.inverse_transform([cnn.predict(scaled_pre_x).flatten()])))
         pre_data['power_forecast'] = res[:len(pre_data)]
         pre_data['farm_id'] = 'J00083'
-        pre_data['cdq'] = 1
-        pre_data['dq'] = 1
-        pre_data['zq'] = 1
+        pre_data['cdq'] = args.get('cdq', 1)
+        pre_data['dq'] = args.get('dq', 1)
+        pre_data['zq'] = args.get('zq', 1)
+        res_cols = ['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']
+        res_cols += [args['target']] if args['algorithm_test'] else res_cols
         pre_data.rename(columns={args['col_time']: 'date_time'}, inplace=True)
-        pre_data = pre_data[['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']]
+        pre_data = pre_data[res_cols]
 
         pre_data['power_forecast'] = pre_data['power_forecast'].round(2)
         pre_data.loc[pre_data['power_forecast'] > g.opt.cap, 'power_forecast'] = g.opt.cap

models_processing/model_koi/tf_cnn_train.py

@@ -18,7 +18,6 @@ from common.logs import Log
 # logger = logging.getLogger()
 logger = Log('tf_cnn').logger
 np.random.seed(42)  # NumPy随机种子
-# tf.set_random_seed(42)  # TensorFlow随机种子
 app = Flask('tf_cnn_train——service')
 
 with app.app_context():
@@ -27,7 +26,6 @@ with app.app_context():
 
     dh = DataHandler(logger, args)
     cnn = CNNHandler(logger, args)
-    global opt
 
 @app.before_request
 def update_config():

models_processing/model_koi/tf_lstm_pre.py

@@ -59,11 +59,13 @@ def model_prediction_bp():
         res = list(chain.from_iterable(target_scaler.inverse_transform([ts.predict(scaled_pre_x).flatten()])))
         pre_data['power_forecast'] = res[:len(pre_data)]
         pre_data['farm_id'] = 'J00083'
-        pre_data['cdq'] = 1
-        pre_data['dq'] = 1
-        pre_data['zq'] = 1
+        pre_data['cdq'] = args.get('cdq', 1)
+        pre_data['dq'] = args.get('dq', 1)
+        pre_data['zq'] = args.get('zq', 1)
+        res_cols = ['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']
+        res_cols += [args['target']] if args['algorithm_test'] else res_cols
         pre_data.rename(columns={args['col_time']: 'date_time'}, inplace=True)
-        pre_data = pre_data[['date_time', 'power_forecast', 'farm_id', 'cdq', 'dq', 'zq']]
+        pre_data = pre_data[res_cols]
 
         pre_data['power_forecast'] = pre_data['power_forecast'].round(2)
         pre_data.loc[pre_data['power_forecast'] > g.opt.cap, 'power_forecast'] = g.opt.cap

models_processing/model_koi/tf_lstm_train.py

@@ -16,7 +16,6 @@ from common.database_dml_koi import *
 from common.logs import Log
 logger = Log('tf_ts').logger
 np.random.seed(42)  # NumPy随机种子
-# tf.set_random_seed(42)  # TensorFlow随机种子
 app = Flask('tf_lstm_train——service')
 
 with app.app_context():
@@ -25,7 +24,6 @@ with app.app_context():
 
     dh = DataHandler(logger, args)
     ts = TSHandler(logger, args)
-    global opt
 
 @app.before_request
 def update_config():