Update app.py

app.py

@@ -1,13 +1,82 @@
 import streamlit as st
 import pandas as pd
 import pickle
+from utils import create_new_features, normalize
+
+min_dict = {'bedrooms': 0,
+ 'bathrooms': 0,
+ 'sqft_living': 370,
+ 'sqft_lot': 638,
+ 'floors': 1,
+ 'waterfront': 0,
+ 'view': 0,
+ 'condition': 1,
+ 'sqft_above': 370,
+ 'sqft_basement': 0,
+ 'yr_built': 1900,
+ 'yr_renovated': 0,
+ 'house_age': 0,
+ 'years_since_renovation': 0}
+
+max_dict = {'bedrooms': 9,
+ 'bathrooms': 8,
+ 'sqft_living': 13540,
+ 'sqft_lot': 1074218,
+ 'floors': 3,
+ 'waterfront': 1,
+ 'view': 4,
+ 'condition': 5,
+ 'sqft_above': 9410,
+ 'sqft_basement': 4820,
+ 'yr_built': 2014,
+ 'yr_renovated': 2014,
+ 'house_age': 114,
+ 'years_since_renovation': 2014}
+
+columns = ['bedrooms', 'bathrooms', 'sqft_living', 'sqft_lot', 'floors',
+       'waterfront', 'view', 'condition', 'sqft_above', 'sqft_basement',
+       'yr_built', 'yr_renovated', 'house_age', 'years_since_renovation',
+       'has_basement', 'city_Algona', 'city_Auburn', 'city_Beaux Arts Village',
+       'city_Bellevue', 'city_Black Diamond', 'city_Bothell', 'city_Burien',
+       'city_Carnation', 'city_Clyde Hill', 'city_Covington',
+       'city_Des Moines', 'city_Duvall', 'city_Enumclaw', 'city_Fall City',
+       'city_Federal Way', 'city_Inglewood-Finn Hill', 'city_Issaquah',
+       'city_Kenmore', 'city_Kent', 'city_Kirkland', 'city_Lake Forest Park',
+       'city_Maple Valley', 'city_Medina', 'city_Mercer Island', 'city_Milton',
+       'city_Newcastle', 'city_Normandy Park', 'city_North Bend',
+       'city_Pacific', 'city_Preston', 'city_Ravensdale', 'city_Redmond',
+       'city_Renton', 'city_Sammamish', 'city_SeaTac', 'city_Seattle',
+       'city_Shoreline', 'city_Skykomish', 'city_Snoqualmie',
+       'city_Snoqualmie Pass', 'city_Tukwila', 'city_Vashon',
+       'city_Woodinville', 'city_Yarrow Point']
 
 with open('./trained_model.pkl', 'rb') as file:
-    loaded_model = pickle.load(file)
+    model = pickle.load(file)
 
 new_pred = st.text_area('Enter text')
 
 if new_pred:
-    new_pred_processed = generate_features(pd.DataFrame([new_pred]), prod=True)
-    out = loaded_model(new_pred_processed)
-    st.json(out[0][0])
+    new_pred = {key:0 for key in X_train.columns}
+    new_pred['date'] = pd.to_datetime('2014-07-10') # do not change
+    
+    new_pred['bedrooms'] = 5
+    new_pred['bathrooms'] = 3
+    new_pred['sqft_living'] = 10000
+    new_pred['sqft_lot'] = 1000
+    new_pred['floors'] = 2
+    new_pred['waterfront'] = 1
+    new_pred['view'] = 3
+    new_pred['condition'] = 5
+    new_pred['sqft_above'] = 500
+    new_pred['sqft_basement'] = 500
+    new_pred['yr_built'] = 2012
+    new_pred['yr_renovated'] = 2013
+    new_pred['city_Bellevue'] = 1
+    new_pred = pd.DataFrame([new_pred])
+    
+    new_pred = create_new_features(new_pred)
+    for col in numerical_features:
+        new_pred[col] = normalize(new_pred, col, min_dict, max_dict)
+    
+    predicted_price = model.predict(new_pred)
+    st.json(predicted_price[0][0])