Create app.py

app.py

@@ -0,0 +1,71 @@
+import streamlit as st
+import pandas as pd
+import joblib
+import pandas as pd
+from sklearn.preprocessing import LabelEncoder, PolynomialFeatures
+
+def preprocess_input(input_data):
+    # Convert input_data to DataFrame if it's not already
+    if not isinstance(input_data, pd.DataFrame):
+        input_data = pd.DataFrame([input_data])
+    
+    # Label encoding for 'cbwd'
+    label_encoder = LabelEncoder()
+    input_data['cbwd'] = label_encoder.fit_transform(input_data['cbwd'])
+    
+    # Feature engineering
+    input_data['season'] = input_data['month'].apply(lambda x: (x % 12 + 3) // 3)
+    input_data['day_of_week'] = pd.to_datetime(input_data[['year', 'month', 'day']]).dt.dayofweek
+    input_data['is_weekend'] = input_data['day_of_week'].apply(lambda x: 1 if x >= 5 else 0)
+    input_data['TEMP_Iws'] = input_data['TEMP'] * input_data['Iws']
+    input_data['DEWP_PRES'] = input_data['DEWP'] * input_data['PRES']
+    
+    # Polynomial features
+    poly = PolynomialFeatures(degree=2, include_bias=False)
+    poly_features = poly.fit_transform(input_data[['DEWP', 'TEMP', 'PRES', 'Iws']])
+    poly_feature_names = poly.get_feature_names_out(['DEWP', 'TEMP', 'PRES', 'Iws'])
+    poly_df = pd.DataFrame(poly_features, columns=poly_feature_names, index=input_data.index)
+    
+    input_data = pd.concat([input_data, poly_df], axis=1)
+    
+    # Select features for prediction
+    features = ['year', 'month', 'day', 'hour', 'DEWP', 'TEMP', 'PRES', 'cbwd', 'Iws', 'Is', 'Ir',
+                'season', 'day_of_week', 'is_weekend', 'TEMP_Iws', 'DEWP_PRES'] + list(poly_feature_names)
+    
+    return input_data[features]
+
+# Load the saved model
+model = joblib.load('random_forest_predictor_pipeline_model.pkl')
+
+st.title('PM2.5 Prediction App')
+
+# Create input fields for all required features
+year = st.number_input('Year', min_value=2000, max_value=2050, value=2024)
+month = st.number_input('Month', min_value=1, max_value=12, value=1)
+day = st.number_input('Day', min_value=1, max_value=31, value=1)
+hour = st.number_input('Hour', min_value=0, max_value=23, value=0)
+DEWP = st.number_input('DEWP', value=0.0)
+TEMP = st.number_input('TEMP', value=0.0)
+PRES = st.number_input('PRES', value=1000.0)
+cbwd = st.selectbox('cbwd', ['NE', 'SE', 'NW', 'cv'])
+Iws = st.number_input('Iws', value=0.0)
+Is = st.number_input('Is', value=0.0)
+Ir = st.number_input('Ir', value=0.0)
+
+if st.button('Predict PM2.5'):
+    # Create a dictionary with the input values
+    input_data = {
+        'year': year, 'month': month, 'day': day, 'hour': hour,
+        'DEWP': DEWP, 'TEMP': TEMP, 'PRES': PRES, 'cbwd': cbwd,
+        'Iws': Iws, 'Is': Is, 'Ir': Ir
+    }
+    
+    # Preprocess the input data
+    processed_input = preprocess_input(input_data)
+    
+    # Make prediction
+    prediction = model.predict(processed_input)
+    
+    st.success(f'The predicted PM2.5 value is: {prediction[0]:.2f}')
+
+