Time Series Forecasting with LightGBM and Prophet | by Gen. David L. | Dec, 2023 | Medium

def prophet_features(df, horizon=24*7): 
    temp_df = df.reset_index() 
    temp_df = temp_df[['datetime', 'count']] 
    temp_df.rename(columns={'datetime': 'ds', 'count': 'y'}, inplace=True) 
 
    # Using the data from the previous week as an example for validation 
    train, test = temp_df.iloc[:-horizon,:], temp_df.iloc[-horizon:,:] 
 
    # Define the Prophet model 
    m = Prophet( 
                growth='linear', 
                seasonality_mode='additive', 
                interval_width=0.95, 
                daily_seasonality=True, 
                weekly_seasonality=True, 
                yearly_seasonality=False 
            ) 
    # Train the Prophet model 
    m.fit(train) 
 
    # Extract features from the data, using Prophet to predict the training set
    predictions_train = m.predict(train.drop('y', axis=1)) 
    # Use Prophet to extract features from the data to predict the test set
    predictions_test = m.predict(test.drop('y', axis=1)) 
    # Combine predictions from the training and test sets
    predictions = pd.concat([predictions_train, predictions_test], axis=0) 
 
    return predictions



def train_time_series_with_folds_autoreg_prophet_features(df, horizon=24*7, lags=[1, 2, 3, 4, 5]): 
    
    # Create a dataframe containing all the new features created with Prophet
    new_prophet_features = prophet_features(df, horizon=horizon) 
    df.reset_index(inplace=True) 
 
    # Merge the Prophet features dataframe with our initial dataframe
    df = pd.merge(df, new_prophet_features, left_on=['datetime'], right_on=['ds'], how='inner') 
    df.drop('ds', axis=1, inplace=True) 
    df.set_index('datetime', inplace=True) 
 
    # Use Prophet predictions to create some lag variables (yhat column)
    for lag in lags: 
        df[f'yhat_lag_{lag}'] = df['yhat'].shift(lag) 
    df.dropna(axis=0, how='any') 
 
    X = df.drop('count', axis=1) 
    y = df['count'] 
 
    # Using the data from the previous week as an example for validation
    X_train, X_test = X.iloc[:-horizon,:], X.iloc[-horizon:,:] 
    y_train, y_test = y.iloc[:-horizon], y.iloc[-horizon:] 
 
    # Define the LightGBM model, train, and make predictions
    model = LGBMRegressor(random_state=42) 
    model.fit(X_train, y_train) 
    predictions = model.predict(X_test) 
 
    # Calculate MAE 
    mae = np.round(mean_absolute_error(y_test, predictions), 3)     
 
    # Plot the real vs prediction for the last week of the dataset
    fig = plt.figure(figsize=(16,6)) 
    plt.title(f'Real vs Prediction - MAE {mae}', fontsize=20) 
    plt.plot(y_test, color='red') 
    plt.plot(pd.Series(predictions, index=y_test.index), color='green') 
    plt.xlabel('Hour', fontsize=16) 
    plt.ylabel('Number of Shared Bikes', fontsize=16) 
    plt.legend(labels=['Real', 'Prediction'], fontsize=16) 
    plt.grid() 
    plt.show()