데이터 분석 과정 작성

server/src/data_processing/main.py

@@ -6,6 +6,7 @@
 """
 
 import datetime
+from server.src.data_processing.model import modeling
 import sys
 import psycopg2
 
@@ -37,15 +38,15 @@ for user in users:
     host = {"email": user[0], "loc_code": user[1], "using_aircon": user[2]}
 
     # 데이터 전처리
-    preprocess(cursor, host)
+    standard_df, mean_df , std_df = preprocess(cursor, host)
 
     # 데이터 분석
-    params = "main_proceduer_function"
+    modeling(standard_df)
 
     # 데이터 분석 결과 저장
-    cursor.execute("INSERT INTO \"Data_Processings\" (host,collected_at,params) VALUES (%s,%s,%s)",
-                   (host["email"], collected_at, params))
+    # cursor.execute("INSERT INTO \"Data_Processings\" (host,collected_at,params) VALUES (%s,%s,%s)",
+    #                (host["email"], collected_at, params))
 
 # Cursor와 Connection 종료
 cursor.close()
-connection.close()
+connection.close()
\ No newline at end of file

server/src/data_processing/model.py

+from preprocessing import standard_df
+import numpy as np
+import tensorflow as tf
+
+def modeling(standard_df):
+    
+    n= len(standard_df)
+    test_size = int(0.3 * n)
+    train = standard_df[:-test_size]
+    test = standard_df[-test_size:]
+    
+    def make_dataset(data, label, window_size = 24):
+        feature_list = []
+        label_list = []
+        for i in range(len(data) - window_size):
+            feature_list.append(np.array(data.iloc[i:i+window_size]))
+            label_list.append(np.array(label.iloc[i + window_size ]))
+        return np.array(feature_list), np.array(label_list)
+    
+    feature_cols = ['temp_out','humi_out','press','wind_speed','Day sin','Day cos','Year sin','Year cos']
+    label_cols = ['temp_out']
+    
+    train_feature = train[feature_cols]
+    train_label = train[label_cols]
+    
+    test_feature = test[feature_cols]
+    test_label = test[label_cols]
+    
+    train_feature, train_label = make_dataset(train_feature, train_label,window_size=6)
+    test_feature, test_label = make_dataset(test_feature, test_label,window_size=6)
+    
+    model = tf.keras.Sequential([
+        tf.keras.layers.LSTM(16,                               
+                            return_sequences=False, 
+                            input_shape = (6,7)),
+        tf.keras.layers.Dense(1)
+                              
+    ])
+    
+    model.compile(loss = 'mse',optimizer = 'adam')
+    model.fit(train_feature, train_label, 
+                        epochs =50, batch_size = 1000)
+    
+    model.save('/src/dataprocessing/model.h5')
+
+
+#사용할때
+# new = tf.keras.models.load_model('/src/dataprocessing/model.h5')
\ No newline at end of file

server/src/data_processing/preprocessing.py

@@ -6,6 +6,10 @@
 
 """
 
+import pandas as pd
+import datetime
+import numpy as np
+
 def preprocess(cursor, host):
     """
         ### preprocess(cursor, host)
@@ -34,3 +38,27 @@ def preprocess(cursor, host):
             result[0], result[1], result[2], result[3], result[4], result[5], result[6], result[7]))
 
     file.close()
+
+
+    df = pd.read_csv("/src/dataprocessing/temp.csv")
+    date_time = pd.to_datetime(df['date'], format='%Y-%m-%d %H:%M')
+    timestamp_s = date_time.map(datetime.datetime.timestamp)
+
+    df = df[['temp_out','humi_out','press','wind_speed']]
+    day = 24*60*60
+    year = (365.2425)*day
+
+    df['Day sin'] = np.sin(timestamp_s * (2 * np.pi / day))
+    df['Day cos'] = np.cos(timestamp_s * (2 * np.pi / day))
+    df['Year sin'] = np.sin(timestamp_s * (2 * np.pi / year))
+    df['Year cos'] = np.cos(timestamp_s * (2 * np.pi / year))
+
+    def standard(dataframe):
+        mean = dataframe.mean()
+        std = dataframe.std()
+        zscore = ( dataframe - mean ) / std
+        return zscore, mean, std
+        
+    standard_df, mean_df , std_df = standard(df)
+
+    return standard_df, mean_df , std_df
\ No newline at end of file