params.yaml

data_ingestion:
  test_size: 0.20
  data_url: './data/Stock Market Dataset.csv'


data_visualization:
  file_path: "./data/transformed_data.csv"  # Update with the actual path
  stock_names:  # List of stock names for which you want to visualize distribution and scatter plots
    - "Apple"
    - "Tesla"
    - "Microsoft"
    - "Google"
  correlation_pairs:  # List of stock pairs or one stock vs list of others to analyze correlation
    - ["Apple", "Tesla"]
    - ["Google", "Microsoft"]
    - ["Apple", "Google", "Tesla"]  # Apple vs Google and Tesla



feature_engineering:
  file_path: './data/transformed_data.csv'  # Path to the raw data CSV file
  output_path: './data/engineered_data.csv'  # Path to save the engineered data CSV file
  missing_value_method: 'fill_mean'  # Method to handle missing values: 'drop', 'fill_mean', or 'fill_median'
  lags: [1, 2, 3]  # List of lags for creating lag features
  windows: [5, 10]  # List of windows for creating rolling statistics

 # Windows for rolling statistics



model_building:
  data_file: 'data/engineered_data.csv'
  save_path: './data'
  model_name: 'RandomForestRegressor'
  hyperparameters:
    n_estimators: 100
    max_depth: 10
    random_state: 42
  model_path: 'models/stock_price_model.pkl'
  predictions_file: 'predictions/stock_price_predictions.csv'
  test_size: 0.2




model_evaluation:
  X_test_path: './data/X_test.csv'
  y_test_path: './data/y_test.csv'
  X_train_path: './data/X_train.csv'
  y_train_path: './data/y_train.csv'
  model_path: './models/stock_price_model.pkl'


prediction:
  stock_name: 'Bitcoin'
  prediction_period: 30  # Number of days to predict
  model_path: 'models/stock_price_model.pkl'
  output_file: 'predictions/apple_predictions.csv'
  X_test_path: './data/X_test.csv'