added coefficient multiplication logic

app.py

+import pandas as pd
+
 if __name__ == "__main__":
     from dotenv import load_dotenv
     load_dotenv(dotenv_path='config.env')
@@ -18,22 +20,22 @@ start_date, end_date, start_timestamp, end_timestamp = KairosStartEndDate().star
 
 def get_tag_details():
     try:
-        df_raw_tags, df_predicted_tags = get_raw_predicted_tags()
+        df_raw_tags, df_predicted_tags, df_coefficients = get_raw_predicted_tags()
         logger.info(f'raw tags dataframe shape - {df_raw_tags.shape}')
         logger.info(f'predicted tags dataframe shape - {df_predicted_tags.shape}')
 
-        df = get_tags_data(df_input_tags=df_raw_tags, start_timestamp=start_timestamp, end_timestamp=end_timestamp)
-        logger.info(f'Shape of final df - {df.shape}')
+        df_kairos_data = get_tags_data(df_input_tags=df_raw_tags, start_timestamp=start_timestamp, end_timestamp=end_timestamp)
+        logger.info(f'Shape of final df - {df_kairos_data.shape}')
 
         mppt_data = GetData()
-        df_mppt = mppt_data.current_voltage_mppt_data(df=df)
-
+        df_mppt = mppt_data.associate_inv_mppt_id(df=df_kairos_data)
+        df = mppt_data.multiply_mppt_coefficients(df_mppt=df_mppt, df_coefficients=df_coefficients)
         data_preprocessing = DataPreprocessing()
-        df_mppt = data_preprocessing.remove_outliers(df=df_mppt, param_list=['tilt_irradiance', 'voltage_mppt',
+        df = data_preprocessing.remove_outliers(df=df, param_list=['tilt_irradiance', 'voltage_mppt',
                                                                    'current_mppt'])
 
-        df_mppt, df_train, df_test = data_preprocessing.train_test_split(df=df_mppt)
-        get_training_inference = TrainingInference(df=df_mppt, df_train=df_train, df_test=df_test)
+        df, df_train, df_test = data_preprocessing.train_test_split(df=df)
+        get_training_inference = TrainingInference(df=df, df_train=df_train, df_test=df_test)
         ai_modelling(df_train=df_train, get_training_inference=get_training_inference,
                      df_predicted_tags=df_predicted_tags)
     except Exception as e:

scripts/core/engine/final_predictions.py

@@ -24,6 +24,12 @@ def ai_modelling(df_train, get_training_inference, df_predicted_tags):
 
                     df_result["timestamp"] = df_result["datetime"].values.astype(np.int64) / 10 ** 9
                     df_result["timestamp"] = df_result["timestamp"].astype('int')
+                    df_result['hour'] = df_result['datetime'].dt.hour
+                    df_result.loc[df_result['hour'].between(left=18, right=23, inclusive=True),
+                                  'predicted_current_mppt'] = 0
+                    df_result.loc[df_result['hour'].between(left=0, right=5, inclusive=True),
+                                  'predicted_current_mppt'] = 0
+                    df_result.drop(['hour'], axis=1, inplace=True)
                     CalculatedDataPush(df_result=df_result, final_tags_dict=final_dict).kafka_data_push()
                     logger.info(f'{final_dict}')
                 except Exception as e:

scripts/core/engine/mppt_data.py

@@ -4,7 +4,7 @@ from loguru import logger
 
 class GetData:
     @staticmethod
-    def current_voltage_mppt_data(df):
+    def associate_inv_mppt_id(df):
         try:
             current_column_list = [col_name for col_name in df if 'current' in col_name]
             voltage_column_list = [col_name for col_name in df if 'voltage' in col_name]
@@ -25,10 +25,24 @@ class GetData:
                     df_mppt = pd.concat([df_mppt, df_temp], axis=0)
                     df_mppt.reset_index(drop=True, inplace=True)
             df_mppt.sort_values(['inv_id', 'mppt_id'], inplace=True)
+            df_mppt['inv_id_mppt_id'] = df_mppt['inv_id'] + '_' + df_mppt['mppt_id']
+            df_mppt.dropna(axis=0, inplace=True)
             df_mppt.reset_index(drop=True, inplace=True)
             return df_mppt
         except Exception as e:
             logger.exception(f'Exception - {e}')
 
+    @staticmethod
+    def multiply_mppt_coefficients(df_mppt, df_coefficients):
+        try:
+            df = pd.merge(df_mppt, df_coefficients, on='inv_id_mppt_id', how='left')
+            logger.debug(f'null rows after mppt coefficient multiplication - {df.isnull().sum()}')
+            df["coefficient"].fillna(1, inplace=True)
+            df.reset_index(drop=True, inplace=True)
+            df['current_mppt'] = df['current_mppt'] * df["coefficient"]
+            df.drop(['coefficient', 'inv_id_mppt_id'], axis=1, inplace=True)
+            return df
+        except Exception as e:
+            logger.exception(f'Exception - {e}')
 
 

scripts/core/engine/raw_predicted_tags.py

@@ -32,6 +32,15 @@ def get_raw_predicted_tags():
             df_raw_tags.rename(columns={'index': 'tag_name'}, inplace=True)
             df_predicted_tags.reset_index(inplace=True)
             df_predicted_tags.rename(columns={'index': 'tag_name'}, inplace=True)
-            return df_raw_tags, df_predicted_tags
+
+            mppt_coefficients = mongo_conn.find_one({"$and": [{"id": "dalmia_string_level_tags"}, {"city": "ariyalur"},
+                                                              {"tags_property": "mppt_coefficients"}]})
+
+            coefficients_dict = mppt_coefficients['input_data']
+            logger.info(f'req coefficients dict length - {len(coefficients_dict)}')
+            df_coefficients = pd.DataFrame.from_dict(coefficients_dict, orient='index')
+            df_coefficients.reset_index(inplace=True)
+            df_coefficients.rename(columns={'index': 'inv_id_mppt_id'}, inplace=True)
+            return df_raw_tags, df_predicted_tags, df_coefficients
     except Exception as e:
         logger.exception(f'Exception - {e}')