🥑 Avocado Price Forecasting System

A production-ready Time Series application that forecasts the average price of conventional avocados in the US using SARIMA.

---

⭐ Features

• Fully automated time-series pipeline
• Modular project architecture
• Statistical validation using ADF Test
• Automatic visualization generation
• Forecast evaluation with multiple metrics
• Production-style repository structure

---

📖 Project Overview

This project analyzes historical avocado prices (2015–2018) to predict future market trends using SARIMA time-series modeling.

The project is structured as a modular Python application, not a notebook.

🎯 Objectives

• Analyze market trend and seasonality
• Test stationarity using the ADF test
• Train a SARIMA forecasting model
• Generate automated visual reports

---

📂 Project Structure

```text
avocado-sales/
│
├── data/
│   └── avocado.csv          # Raw dataset (Kaggle)
│
├── src/                     # Application Source Code
│   ├── main.py              # 🚀 Entry point
│   ├── loader.py            # Data cleaning & preprocessing
│   ├── model.py             # Statistical modeling (SARIMA)
│   ├── plots.py             # Visualization engine
│   └── pics/                # Generated charts (saved automatically)
│
├
├── requirements.txt
└── README.md

---

🚀 How to Run

Install dependencies

pip install -r requirements.txt

Run the pipeline

cd src
python main.py

The script will:

• Load data
• Train SARIMA model
• Print metrics
• Generate charts automatically

---

📊 Model Performance

Metric	Score	Interpretation
RMSE	0.22	±22 cents error
MAE	0.18	Avg absolute error
MAPE	13.58%	~86.5% accuracy

---

🛠 Tech Stack

• Pandas
• Statsmodels
• Matplotlib / Seaborn
• Scikit-Learn

---

💡 Future Improvements

• Prophet & LSTM comparison
• Streamlit deployment
• Auto hyperparameter tuning

---

👩‍💻 Author

Konul Jafarova