Update app.py

app.py

@@ -2,72 +2,90 @@ import streamlit as st
 import hopsworks
 import joblib
 import pandas as pd
-import datetime
+from datetime import timedelta, datetime
 from functions import *
-import pytz
 
-st.set_page_config(layout="wide")
-
-st.title('AQI prediction for Beijing in next week')
-
-project = hopsworks.login()
 
-#today=datetime.datetime.now(pytz.timezone('Asia/Shanghai')).date()
-#today=datetime.date.today()
+def fancy_header(text, font_size=24):
+    res = f'<p style="color:#ff5f72; font-size: {font_size}px; text-align:center;">{text}</p>'
+    st.markdown(res, unsafe_allow_html=True)
 
-#weekly_data = get_weather_data_weekly(today)
-
-# get Hopsworks Model Registry
-#mr = project.get_model_registry()
-# get model object
-#model = mr.get_model("gradient_boost_model", version=1)
-#model_dir = model.download()
-#model = joblib.load(model_dir + "/model.pkl")
+st.set_page_config(layout="wide")
 
-#weekly_data = data_encoder(weekly_data)
+st.title('Air Quality Prediction Project🌩')
 
-#start test
+st.write(9 * "-")
+fancy_header('\n Connecting to Hopsworks Feature Store...')
 
-fs = project.get_feature_store()
-feature_view = fs.get_feature_view(
-    name = 'hel_air_fv1',
-    version = 1
-)
+project = hopsworks.login()
 
-#start_date = datetime.now() - timedelta(days=1)
-#start_time = int(start_date.timestamp()) * 1000
+st.write("Successfully connected!✔️")
 
-#end_date = datetime.now() - timedelta(days=1)
-#end_time = int(start_date.timestamp()) * 1000
+st.write(18 * "-")
+fancy_header('\n Getting data from Feature Store...')
 
-#print("start time:", start_time)
+today = datetime.date.today()
+weekly_data = get_weather_data_weekly(today)
 
-X = feature_view.get_batch_data(start_time=1670194800000)
 
-print("batch data")
-print(X.tail(10))
+st.write(27 * "-")
 
-X = X.drop(columns=["date"]).fillna(0)
+mr = project.get_model_registry()
+# get Hopsworks Model Registry
+mr = project.get_model_registry()
+# get model object
+model = mr.get_model("aqi_model_gb", version=1)
+model_dir = model.download()
+model = joblib.load(model_dir + "/gb_model.pkl")
+
+weekly_data['aqi'] = 0
+weekly_data['city'] = 0
+
+weekly_data = data_encoder(weekly_data)
+
+weekly_data.drop(['tempmax'], inplace = True, axis = 1)
+weekly_data.drop('tempmin', inplace = True, axis = 1)
+weekly_data.drop('feelslikemax', inplace = True, axis = 1)
+weekly_data.drop('feelslikemin', inplace = True, axis = 1)
+weekly_data.drop('feelslike', inplace = True, axis = 1)
+weekly_data.drop('dew', inplace = True, axis = 1)
+weekly_data.drop('precipprob', inplace = True, axis = 1)
+weekly_data.drop('precipcover', inplace = True, axis = 1)
+weekly_data.drop('snow', inplace = True, axis = 1)
+weekly_data.drop('snowdepth', inplace = True, axis = 1)
+weekly_data.drop('windgust', inplace = True, axis = 1)
+weekly_data.drop('windspeed', inplace = True, axis = 1)
+weekly_data.drop('winddir', inplace = True, axis = 1)
+weekly_data.drop('solarradiation', inplace = True, axis = 1)
+weekly_data.drop('solarenergy', inplace = True, axis = 1)
+weekly_data.drop('pressure', inplace = True, axis = 1)
+
+print(weekly_data)
+
+preds=model.predict(weekly_data)
+#model = get_model(project=project,
+#                      model_name="gradient_boost_model",
+#                      evaluation_metric="f1_score",
+#                      sort_metrics_by="max")
+#print("here")
+#print(model)
+#model_dir = model.download()
+#model = joblib.load(model_dir + "/model.pkl")
 
-data_to_display = decode_features(X, feature_view=feature_view)
+st.write("-" * 36)
 
-preds=model.predict(X)
 
-#preds=model.predict(weekly_data)
+#preds = model.predict(data_encoder(weekly_data)).astype(int)
+poll_level = get_aplevel(preds.T.reshape(-1, 1))
 
 next_week = [f"{(today + timedelta(days=d)).strftime('%Y-%m-%d')},{(today + timedelta(days=d)).strftime('%A')}" for d in range(7)]
 
-print("preds")
-print(preds)
-
-aqi_level = encoder_range(preds.T.reshape(-1, 1))
-#df = pd.DataFrame(data=[map(int,preds), aqi_level], index=["aqi","Air Pollution Level"], columns=next_week)
+print(next_week)
 
-print("aqi")
-print(aqi_level)
+df = pd.DataFrame(data=[preds, poll_level], index=["AQI", "Air pollution level"], columns=next_week)
 
-df = pd.DataFrame(data=[map(int,preds), aqi_level], index=["aqi","Air Pollution Level"],columns=next_week)
+st.write(df)
 
-st.write(df) 
+print(st)
 
-st.button("Re-run")
+#st.button("Re-run")