Redeploy

app.py

@@ -1,27 +1,27 @@
 import streamlit as st
 import pandas as pd
-import numpy as np
 import joblib
+import numpy as np
 import seaborn as sns
 import matplotlib.pyplot as plt
-import json
 
-# Fungsi untuk memuat data
+# Function to load data
 def load_data():
     df = pd.read_csv('h8dsft_P1G3_Muhammad_Arief_Kurniawan.csv')
     return df
 
-# Fungsi untuk memuat komponen model dan preprocessing
+# Function to load model and preprocessing components
 def load_components():
-    model = joblib.load('stacked_es_model.pkl')
+    rf_model = joblib.load('rf_model.pkl')
+    gb_model = joblib.load('gb_model.pkl')
+    stacked_model = joblib.load('stacked_es_model.pkl')
     scaler = joblib.load('model_scaler.pkl')
-    with open('numeric_columns.txt', 'r') as file_1:
-        numeric_columns = json.load(file_1)
-    with open('categorical_columns.txt', 'r') as file_2:
-        categorical_columns = json.load(file_2)
-    return model, scaler, numeric_columns, categorical_columns
+    numeric_columns = ['age', 'creatinine_phosphokinase', 'ejection_fraction', 'platelets', 'serum_creatinine', 'serum_sodium', 'time']
+    categorical_columns = ['anaemia', 'diabetes', 'high_blood_pressure', 'sex', 'smoking']
+    trained_features = rf_model.feature_names_in_
+    return rf_model, gb_model, stacked_model, scaler, numeric_columns, categorical_columns, trained_features
 
-# Fungsi untuk mendapatkan input dari pengguna
+# Function to get user input
 def get_user_input(numeric_columns, categorical_columns):
     inputs = {}
     for col in numeric_columns + categorical_columns:
@@ -30,12 +30,14 @@ def get_user_input(numeric_columns, categorical_columns):
         else:
             default_val = np.expm1(50.0) if col != 'time' else 50.0
             inputs[col] = np.log1p(st.number_input(col, value=default_val))
-    return pd.DataFrame([inputs])
+    user_data = pd.DataFrame([inputs])
+    user_data = user_data[numeric_columns + categorical_columns]
+    return user_data
 
 def main():
     st.title("Prediksi dan Eksplorasi Gagal Jantung")
 
-    # Sidebar untuk navigasi
+    # Sidebar for navigation
     choice = st.sidebar.selectbox("Pilih Halaman", ["Beranda", "Eksplorasi Data", "Prediksi"])
 
     if choice == "Beranda":
@@ -57,14 +59,20 @@ def main():
         st.pyplot(fig)
 
     elif choice == "Prediksi":
-        model, scaler, numeric_columns, categorical_columns = load_components()
+        rf_model, gb_model, stacked_model, scaler, numeric_columns, categorical_columns, trained_features = load_components()
         user_input = get_user_input(numeric_columns, categorical_columns)
-        user_input[numeric_columns] = scaler.transform(user_input[numeric_columns])
-        prediction = model.predict(user_input)
-        if prediction[0] == 0:
-            st.write("Prediksi: Tidak Ada Gagal Jantung")
-        else:
-            st.write("Prediksi: Gagal Jantung")
+        user_input_ordered = user_input[trained_features].to_numpy()
+        user_input_ordered[:, :len(numeric_columns)] = scaler.transform(user_input_ordered[:, :len(numeric_columns)])
+        
+        # Predictions
+        rf_prediction = rf_model.predict(user_input_ordered)
+        gb_prediction = gb_model.predict(user_input_ordered)
+        stacked_prediction = stacked_model.predict(user_input_ordered)
+        
+        st.write(f"Random Forest Prediction: {'Gagal Jantung' if rf_prediction[0] == 1 else 'Tidak Ada Gagal Jantung'}")
+        st.write(f"Gradient Boosting Prediction: {'Gagal Jantung' if gb_prediction[0] == 1 else 'Tidak Ada Gagal Jantung'}")
+        st.write(f"Stacked Model Prediction: {'Gagal Jantung' if stacked_prediction[0] == 1 else 'Tidak Ada Gagal Jantung'}")
+        
         st.subheader("Data Input Pengguna")
         st.write(user_input)
 

gb_model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b56ea99334252e1a2cfac5a4e78adc4f70a76ddc8c5049d3f7597da7457de34
+size 242044

model_scaler.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ccba0dda6850d57aae324ed74dc8a9b7b65d71a5c478c1ddd1ca682eea6c68b
+oid sha256:3cba4d82866e259a46c8b31577581f032e6ea1cad13ff97960e019be10c91200
 size 810

numeric_columns.txt

@@ -1 +1 @@
-["age", "creatinine_phosphokinase", "ejection_fraction", "platelets", "serum_creatinine", "serum_sodium", "time"]
+["age", "creatinine_phosphokinase", "ejection_fraction", "platelets", "serum_creatinine", "serum_sodium", "time"]

prediction.py

@@ -2,56 +2,61 @@ import streamlit as st
 import pandas as pd
 import joblib
 import numpy as np
-import json
 
-# Fungsi untuk memuat model dan komponen preprocessing
 def load_components():
-    model = joblib.load('stacked_es_model.pkl')
+    rf_model = joblib.load('rf_model.pkl')
+    gb_model = joblib.load('gb_model.pkl')
+    stacked_model = joblib.load('stacked_es_model.pkl')
     scaler = joblib.load('model_scaler.pkl')
-    with open('numeric_columns.txt', 'r') as file_1:
-        numeric_columns = json.load(file_1)
-    with open('categorical_columns.txt', 'r') as file_2:
-        categorical_columns = json.load(file_2)
-    return model, scaler, numeric_columns, categorical_columns
+    numeric_columns = ['age', 'creatinine_phosphokinase', 'ejection_fraction', 'platelets', 'serum_creatinine', 'serum_sodium', 'time']
+    categorical_columns = ['anaemia', 'diabetes', 'high_blood_pressure', 'sex', 'smoking']
+    
+    trained_features = rf_model.feature_names_in_  # Extract feature names from the model
+
+    return rf_model, gb_model, stacked_model, scaler, numeric_columns, categorical_columns, trained_features
 
-# Fungsi untuk mendapatkan input dari pengguna
+# Function to get user input
 def get_user_input(numeric_columns, categorical_columns):
     inputs = {}
     for col in numeric_columns + categorical_columns:
         if col in categorical_columns:
             inputs[col] = st.selectbox(col, [0, 1])
         else:
-            # Using np.expm1 to reverse the log transformation for display and input
             default_val = np.expm1(50.0) if col != 'time' else 50.0
             inputs[col] = np.log1p(st.number_input(col, value=default_val))
-    user_df = pd.DataFrame([inputs])
     
-    # Ensure the columns are in the expected order
-    return user_df[numeric_columns + categorical_columns]
-
+    user_data = pd.DataFrame([inputs])
+    user_data = user_data[numeric_columns + categorical_columns]  # Ensure correct column order
+    return user_data
 
 def main():
     st.title("Prediksi Gagal Jantung")
 
-    # Muat model dan komponen preprocessing
-    model, scaler, numeric_columns, categorical_columns = load_components()
+    # Load models and preprocessing components
+    rf_model, gb_model, stacked_model, scaler, numeric_columns, categorical_columns, trained_features = load_components()
 
-    # Dapatkan input dari pengguna
+    # Get user input
     user_input = get_user_input(numeric_columns, categorical_columns)
 
+    # Reorder columns to match trained features and convert to numpy array
+    user_input_ordered = user_input[trained_features].to_numpy()
+
     # Preprocess input
-    user_input[numeric_columns] = scaler.transform(user_input[numeric_columns])
+    user_input_ordered[:, :len(numeric_columns)] = scaler.transform(user_input_ordered[:, :len(numeric_columns)])
+
+    # Make predictions using both models
+    rf_prediction = rf_model.predict(user_input_ordered)
+    gb_prediction = gb_model.predict(user_input_ordered)
+    stacked_prediction = stacked_model.predict(user_input_ordered)
 
-    # Lakukan prediksi
-    prediction = model.predict(user_input)
-    if prediction[0] == 0:
-        st.write("Prediksi: Tidak Ada Gagal Jantung")
-    else:
-        st.write("Prediksi: Gagal Jantung")
+    st.write(f"Random Forest Prediction: {'Gagal Jantung' if rf_prediction[0] == 1 else 'Tidak Ada Gagal Jantung'}")
+    st.write(f"Gradient Boosting Prediction: {'Gagal Jantung' if gb_prediction[0] == 1 else 'Tidak Ada Gagal Jantung'}")
+    st.write(f"stacked Prediction: {'Gagal Jantung' if stacked_prediction [0] == 1 else 'Tidak Ada Gagal Jantung'}")
 
-    # Opsi: Tampilkan data input (untuk verifikasi)
+    # Option: Display user input data (for verification)
     st.subheader("Data Input Pengguna")
     st.write(user_input)
 
 if __name__ == "__main__":
     main()
+

requirements.txt

@@ -3,4 +3,4 @@ pandas
 seaborn
 matplotlib
 numpy
-scikit-learn
+scikit-learn==1.2.2

rf_model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba7553ca39091bcf03b2a38d8ef4753f881828d206d0e8396018e31120e20758
+size 761193

stacked_es_model.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9053ef824aec6ff6f04fc049702d52b829c992e50e661641c38357dad77ed95
-size 641887
+oid sha256:4bfb4266106c8923e3e14fa886edc329eb3dcf8d026b784317d7fb25bfa844a2
+size 632416