completed first version energy prediction pipeline

.gitignore

@@ -5,4 +5,3 @@ __pycache__/
 *.tf
 data
 *.csv
-src/test_main.py

EnergyLSTM/models/scalerNorth.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b6243927a7de9f60f2760a039d9b44c80a8f53662540cdb3a3ac878e671baf6
+size 689

src/energy_prediction/EnergyPredictionPipeline.py

@@ -1,15 +1,83 @@
 import pandas as pd
+from pickle import load
+from datetime import datetime, date
+from sklearn.preprocessing import StandardScaler
+import joblib
+import json
+import numpy as np
 
 class EnergyPredictionPipeline:
-    def __init__(self, model_path_north, model_path_south):
-        self.model_north = EnergyPredictionNorth(model_path_north)
-        self.model_south = EnergyPredictionSouth(model_path_south)
+    scalerNorth = None 
+    scalerSouth = None 
 
-    def predict(self, data):
-        north_data = data[data["wing"] == "north"]
-        south_data = data[data["wing"] == "south"]
+    def __init__(self, scaler1_path=None,scaler2_path=None):
+        
+        if scaler1_path:
+            self.scalerNorth = self.get_scaler(scaler1_path)
+        if scaler2_path:
+            self.scalerSouth = self.get_scaler(scaler2_path)
+        
+        self.input_col_names = self.input_col_names + [
+            "date",
+            "hvac_N"
+            ]
+     
+    def get_scaler(self, scaler_path):
+        return joblib.load(scaler_path)
+    
+    def transform_windows(self, df):
+        return self.scalerNorth.transform(df)
 
-        north_prediction = self.model_north.predict(north_data)
-        south_prediction = self.model_south.predict(south_data)
+    def date_encoder(df):
 
-        return pd.concat([north_prediction, south_prediction])
+        df['day_of_week']  = df.index.dayofweek
+        df['hour_of_day']  = df.index.hour
+        df['month']        = df.index.month
+
+        df['day_encoding'] = np.sin(2*np.pi*df['day_of_week']/7)
+        df['hour_encoding'] = np.sin(2*np.pi*df['hour_of_day']/24)
+        df['month_encoding'] = np.sin(2*np.pi*df['month']/12)
+
+        return df
+    
+    def prepare_input(self, df_new):
+        
+        df = df_new.copy()
+        df["date"] = pd.to_datetime(df["date"])
+        df.set_index("date", inplace=True)
+        df = df.resample("H").mean()
+        
+        df = self.date_encoder(df)
+
+        df.reset_index(inplace=True, drop=True)
+
+        return df
+    
+    def extract_data_from_message(self, message):
+        payload = json.loads(message.payload.decode())
+
+        len_df = len(self.df)
+
+        k = {}
+        for col in self.input_col_names:
+            k[col] = payload[col]
+        self.df.loc[len_df] = k
+        return self.df
+      
+    def get_window(self, df):
+        len_df = len(df)
+        print(len_df)
+        if len_df > 4*7*24:
+            return df[len_df - 673 : len_df].astype("float32")
+        else:
+            return None
+        
+    def fit(self, message):
+        df_new      = self.extract_data_from_message(message)
+        df_window   = self.get_window(df_new)
+        if df_window is not None:
+            df = self.prepare_input(df_window)
+            df = self.transform_windows(df)
+        else:
+            df = None
+        return df

src/energy_prediction/models/scalerNorth.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b6243927a7de9f60f2760a039d9b44c80a8f53662540cdb3a3ac878e671baf6
+size 689

src/test_main.py

@@ -0,0 +1,29 @@
+from energy_prediction.EnergyPredictionNorth import EnergyPredictionNorth
+from energy_prediction.EnergyPredictionSouth import EnergyPredictionSouth
+from energy_prediction.EnergyPredictionPipeline import EnergyPredictionPipeline
+
+def main():
+    # Energy Prediction North wing
+    EnergyPredictionNorth = EnergyPredictionNorth(
+        model_path="src/energy_prediction/models/lstm_energy_north_01.keras"
+    )
+    # Energy Prediction South wing
+
+    def on_message(client, userdata, message):
+        df = EnergyPredictionPipeline.fit(message)
+        
+        if not df is None:
+            out_vav = EnergyPredictionNorth.pipeline(df, EnergyPredictionPipeline.scaler)
+
+    broker_address = "localhost"
+    broker_port = 1883
+    topic = "sensor_data"
+    client = mqtt.Client(mqtt.CallbackAPIVersion.VERSION1)
+    print("Connecting to broker")
+    client.on_message = on_message
+    client.connect(broker_address, broker_port)
+    client.subscribe(topic)
+    client.loop_forever()
+
+if __name__ == "__main__":
+    main()