Update app.py

app.py

@@ -5,9 +5,11 @@ import pandas as pd
 import matplotlib.pyplot as plt
 import seaborn as sns
 import scipy.optimize as sco
-from sklearn.linear_model import LinearRegression
-from sklearn.model_selection import train_test_split
-from sklearn.preprocessing import StandardScaler
+import tensorflow as tf
+from tensorflow import keras
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import LSTM, Dense, Dropout
+from sklearn.preprocessing import MinMaxScaler
 
 def get_stock_data(tickers, start, end):
     data = yf.download(tickers, start=start, end=end)
@@ -60,27 +62,6 @@ def generate_efficient_frontier(returns, cov_matrix, num_portfolios=5000):
     
     return results
 
-def predict_stock_price(data, days=30):
-    stock = st.selectbox("Pilih saham untuk prediksi", data.columns)
-    df = data[[stock]].dropna()
-    df['Days'] = np.arange(len(df))
-    
-    X = df[['Days']]
-    y = df[stock]
-    
-    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
-    model = LinearRegression()
-    model.fit(X_train, y_train)
-    
-    future_days = np.arange(len(df), len(df) + days).reshape(-1, 1)
-    predictions = model.predict(future_days)
-    
-    fig, ax = plt.subplots()
-    ax.plot(df.index, df[stock], label='Harga Aktual')
-    ax.plot(pd.date_range(df.index[-1], periods=days, freq='D'), predictions, label='Prediksi', linestyle='dashed')
-    ax.legend()
-    st.pyplot(fig)
-
 def plot_correlation_heatmap(data):
     correlation_matrix = data.corr()
     fig, ax = plt.subplots()
@@ -103,6 +84,39 @@ def simulate_dca(data, investment_amount=100):
     total_value.plot(ax=ax, title='Simulasi Investasi DCA')
     st.pyplot(fig)
 
+def predict_stock_price(data, stock, days=30):
+    st.subheader(f"Prediksi Harga Saham {stock} (LSTM Model)")
+    scaler = MinMaxScaler(feature_range=(0, 1))
+    scaled_data = scaler.fit_transform(data[stock].values.reshape(-1,1))
+    
+    X, y = [], []
+    for i in range(60, len(scaled_data) - days):
+        X.append(scaled_data[i-60:i, 0])
+        y.append(scaled_data[i:i+days, 0])
+    
+    X, y = np.array(X), np.array(y)
+    X = np.reshape(X, (X.shape[0], X.shape[1], 1))
+    
+    model = Sequential()
+    model.add(LSTM(units=50, return_sequences=True, input_shape=(X.shape[1], 1)))
+    model.add(LSTM(units=50, return_sequences=False))
+    model.add(Dense(units=25))
+    model.add(Dense(units=days))
+    
+    model.compile(optimizer='adam', loss='mean_squared_error')
+    model.fit(X, y, epochs=10, batch_size=16)
+    
+    predicted_prices = model.predict(X[-1].reshape(1, X.shape[1], 1))
+    predicted_prices = scaler.inverse_transform(predicted_prices)
+    
+    future_dates = pd.date_range(data.index[-1], periods=days+1)[1:]
+    
+    fig, ax = plt.subplots()
+    ax.plot(data.index, data[stock], label='Data Historis')
+    ax.plot(future_dates, predicted_prices.flatten(), label='Prediksi', linestyle='dashed')
+    ax.legend()
+    st.pyplot(fig)
+
 st.title("Analisis Portofolio Saham Optimal (Model Markowitz)")
 
 st.write("Rekomendasi Saham yang Bertahan Saat COVID-19:")
@@ -131,21 +145,12 @@ if st.button("Analisis Portofolio"):
         st.subheader("Simulasi Investasi Dollar Cost Averaging (DCA)")
         simulate_dca(stock_data)
         
-        st.subheader("Prediksi Harga Saham")
-        predict_stock_price(stock_data)
+        st.subheader("Prediksi Harga Saham dengan AI")
+        predict_stock_price(stock_data, selected_stock)
         
         if optimal_weights is not None:
             st.subheader("Bobot Portofolio Optimal")
             portfolio_weights = {stock: weight for stock, weight in zip(stock_data.columns, optimal_weights)}
             st.write(portfolio_weights)
-            
-            fig, ax = plt.subplots()
-            ax.pie(portfolio_weights.values(), labels=portfolio_weights.keys(), autopct='%1.1f%%', startangle=140)
-            ax.axis('equal')
-            st.pyplot(fig)
-        else:
-            st.error("Optimasi portofolio gagal. Coba dengan saham yang berbeda.")
-
 
-#Saya telah menambahkan fitur prediksi harga saham menggunakan model Linear Regression. Sekarang, pengguna dapat memilih saham dan melihat prediksi harganya dalam beberapa hari ke depan. Grafik juga telah disertakan untuk membandingkan harga aktual dengan prediksi. Silakan coba dan beri tahu jika ada yang perlu diperbaiki!