first commit

app.py

@@ -0,0 +1,44 @@
+# from src import stock_data, model_train, train_test_split, model_predict 
+# import streamlit as st
+# user_input = st.text_input("Enter your stock ticker here:")
+# train = stock_data (user_input)
+# x_train, y_train, scaler = train_test_split(train)
+# history, model = model_train(x_train, y_train)
+# result, data_inf = model_predict(train, scaler, model) 
+
+import streamlit as st
+from src import stock_data, model_train, train_test_split, model_predict
+
+# Function to display result, history, and data information
+def display_results(user_input,result, history, data_inf):
+    st.header(f'Here is the data for {user_input} the past 10 days.')
+    st.write(data_inf)
+    st.markdown(f'''
+        Stock Prediction Analysis for <span style="font-size:24px;">{user_input}</span>
+        
+        <p style="font-size:24px;">
+        Tomorrow's {user_input} predicted price is : <b>{result}</b>
+        </p>
+    ''', unsafe_allow_html=True)
+    
+    
+    
+# Main function to run the app
+def main():
+    st.title("Stock Prediction App")
+    user_input = st.text_input("Enter your stock ticker here:")
+    
+    if st.button("Predict"):
+        with st.spinner('Loading...'):
+            train = stock_data(user_input)
+            if len(train)>0: 
+                x_train, y_train, scaler = train_test_split(train)
+                history, model = model_train(x_train, y_train)
+                result, data_inf = model_predict(train, scaler, model) 
+                display_results(user_input,result[0][0], history, data_inf)
+            else:
+                st.write('Invalid stock ticker. Please verify the ticker symbol on the following website: [Yahoo Finance.](https://finance.yahoo.com/)')
+
+
+if __name__ == "__main__":
+    main()

requirements.txt

@@ -0,0 +1,5 @@
+scikit-learn==1.3.2
+yfinance==0.2.37
+tensorflow==2.15.0
+numpy==1.26.3
+streamlit

src/__init__.py

@@ -0,0 +1,3 @@
+from .model import *
+from .get_data import *
+from .train_test_split_data import *

src/get_data.py

@@ -0,0 +1,8 @@
+import yfinance as yf
+from datetime import datetime, timedelta
+from sklearn.preprocessing import MinMaxScaler
+import numpy as np
+
+def stock_data (tick):
+  data = yf.download(tick, start=datetime.now() - timedelta(days=365),end=datetime.now())
+  return data.Close

src/model.py

@@ -0,0 +1,35 @@
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import GRU, Bidirectional, Dense, Dropout
+from tensorflow.keras.callbacks import EarlyStopping
+from tensorflow.keras.regularizers import L2
+import tensorflow as tf
+
+def model_train (x_train, y_train):
+  tf.random.set_seed(100)
+  tf.keras.backend.clear_session()
+
+  model = Sequential()
+  model.add(Bidirectional(GRU(units=256, activation='relu', return_sequences=True, input_shape=(10,1))))
+  model.add(Dropout(0.5))
+  model.add(Bidirectional(GRU(units=256, activation='relu', return_sequences=False, kernel_regularizer=L2(0.01))))
+  model.add(Dropout(0.1))
+  model.add(Dense(units=1, activation='linear')) # Prediction of the next value
+
+  model.compile(optimizer='adam', loss='mape', metrics='mape')
+
+  model.build((None, 10, 1))  # Specify input shape
+
+  # Define EarlyStopping callback
+  early_stopping = EarlyStopping(monitor='val_loss', patience=3, restore_best_weights=True)
+
+  # Train the model
+  history = model.fit(x_train, y_train, epochs=10, batch_size=20, validation_split=0.2, callbacks=[early_stopping])
+  return history, model
+
+def model_predict(train,scaler,model):
+  data_inf = train[-10:]
+  scaled_data_inf = scaler.transform(data_inf.values.reshape(-1,1))
+  data_inf_final = tf.expand_dims(scaled_data_inf,0)
+  pred_inf = model.predict(data_inf_final)
+  return scaler.inverse_transform(pred_inf) , data_inf
+

src/train_test_split_data.py

@@ -0,0 +1,20 @@
+from sklearn.preprocessing import MinMaxScaler
+import numpy as np
+
+def train_test_split(train):
+    scaler = MinMaxScaler()
+    scaled_data = scaler.fit_transform(train.values.reshape(-1,1))
+
+    prediction_days = 10
+
+    x_train = []
+    y_train = []
+
+    for x in range(prediction_days, len(scaled_data)):
+        x_train.append(scaled_data[x-prediction_days:x, 0])
+        y_train.append(scaled_data[x, 0])
+
+
+    x_train, y_train = np.array(x_train), np.array(y_train)
+    x_train = np.reshape(x_train, (x_train.shape[0], x_train.shape[1], 1))
+    return x_train, y_train , scaler