3.2 Case study Profit Forecast.ipynb

Kernel: Python 3 (ipykernel)

Profit Forecast using ARIMA Model

Load Data

• How can we load the historical profit data from the CSV file?

• What preprocessing steps are necessary before fitting the ARIMA model?

• How can we determine the optimal ARIMA parameters for the profit data?

• What evaluation metrics can we use to assess the accuracy of the ARIMA model?

• How can we use the ARIMA model to forecast profits for the next 12 months?

• What visualizations can we create to interpret the historical data, model residuals, and forecasted profits?

import pandas as pd
from statsmodels.tsa.arima.model import ARIMA
import matplotlib.pyplot as plt
from statsmodels.graphics.tsaplots import plot_acf, plot_pacf
from statsmodels.tsa.stattools import adfuller
from sklearn.metrics import mean_absolute_error, mean_squared_error
import numpy as np

# Step 1: Problem Statement
print("Case Study: Profit Forecast using ARIMA")

# Step 2: Load Data
def load_data(file_path):
    """
    Load historical profit data from a CSV file into a pandas DataFrame.
    """
    return pd.read_csv(file_path)

# Question: How can we load the historical profit data from the CSV file?

# Step 3: Preprocess Data
def preprocess_data(data):
    """
    Preprocess the profit data:
    - Convert date column to datetime format
    - Check for missing values
    """
    # Convert date column to datetime
    data['Date'] = pd.to_datetime(data['Date'])
    
    # Check for missing values
    if data.isnull().values.any():
        print("Warning: Missing values found in the data.")
    
    return data

# Question: What preprocessing steps are necessary before fitting the ARIMA model?

# Step 4: Fit ARIMA Model
def fit_arima_model(data, p, d, q):
    """
    Fit ARIMA model to the profit data.
    """
    # Fit ARIMA model
    model = ARIMA(data, order=(p, d, q))
    model_fit = model.fit()
    
    return model_fit

# Question: How can we determine the optimal ARIMA parameters for the profit data?

# Step 5: Evaluate Model
def evaluate_model(model_fit, test_data):
    """
    Evaluate ARIMA model's performance using Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE).
    """
    # Make predictions
    forecast = model_fit.forecast(steps=len(test_data))
    
    # Calculate evaluation metrics
    mae = mean_absolute_error(test_data, forecast)
    mse = mean_squared_error(test_data, forecast)
    rmse = np.sqrt(mse)
    
    return mae, mse, rmse

# Question: What evaluation metrics can we use to assess the accuracy of the ARIMA model?

# Step 6: Forecast Future Profits
def forecast_profits(model_fit, periods):
    """
    Forecast future profits using the ARIMA model.
    """
    # Forecast future profits
    forecast = model_fit.forecast(steps=periods)
    
    return forecast

# Question: How can we use the ARIMA model to forecast profits for the next 12 months?

# Step 7: Visualization
def visualize_data(data):
    """
    Visualize historical profit data.
    """
    plt.figure(figsize=(10, 6))
    plt.plot(data['Date'], data['Profit'], label='Historical Profit Data')
    plt.xlabel('Date')
    plt.ylabel('Profit')
    plt.title('Historical Profit Data')
    plt.legend()
    plt.show()

# Question: What visualizations can we create to interpret the historical data, model residuals, and forecasted profits?

# Main Function
def main():
    # Step 2: Load Data
    file_path = 'profit_data.csv'
    profit_data = load_data(file_path)

    # Step 3: Preprocess Data
    profit_data = preprocess_data(profit_data)

    # Step 4: Fit ARIMA Model
    p, d, q = 2, 1, 2  # Example ARIMA parameters
    model_fit = fit_arima_model(profit_data['Profit'], p, d, q)

    # Step 5: Evaluate Model
    test_data = profit_data['Profit'][-12:]  # Example: Using last 12 months as test data
    mae, mse, rmse = evaluate_model(model_fit, test_data)
    print("Evaluation Metrics:")
    print("Mean Absolute Error (MAE):", mae)
    print("Mean Squared Error (MSE):", mse)
    print("Root Mean Squared Error (RMSE):", rmse)

    # Step 6: Forecast Future Profits
    forecast_period = 12  # Example: Forecasting profits for the next 12 months
    forecast = forecast_profits(model_fit, forecast_period)

    # Step 7: Visualization
    visualize_data(profit_data)

# Execute Main Function
if __name__ == "__main__":
    main()

Case Study: Profit Forecast using ARIMA
Evaluation Metrics:
Mean Absolute Error (MAE): 1016.2633640091667
Mean Squared Error (MSE): 1539871.6265948142
Root Mean Squared Error (RMSE): 1240.9156404022049

C:\Users\suyashi144893\Anaconda3\lib\site-packages\statsmodels\tsa\statespace\sarimax.py:978: UserWarning: Non-invertible starting MA parameters found. Using zeros as starting parameters.
  warn('Non-invertible starting MA parameters found.'