bawcode/ethiopian-coffee-trade-analysis

# Ethiopian Coffee Trade Analysis ## Overview This repository contains a comprehensive data science analysis of **44,629 coffee transactions** from the **Ethiopian Commodity Exchange (ECX)**, spanning **January 2018 to March 2021**. The project uncovers critical market dynamics in Ethiopia's coffee trade — the birthplace of coffee — where it accounts for **30–35% of export revenues** and supports **over 15 million livelihoods**. Using Python tools, the analysis reveals: - Strong **negative price-volume correlation** (`r = -0.88`) → bulk discounts - **44% premium** for **Harar coffee** at **2,036.15 ETB/quintal** - **Seasonal peak in February** at **1,501.08 ETB/quintal** - **Gimbi** as top logistics hub handling **1.92M kg** - **Prophet forecast**: Price surge to **2,194 ETB/quintal by June 2021** - Data challenge: **40% "Unknown" grades** causing volatility Developed during an **INSA internship** by **Besufekad Ayalkbet** (Data Science, Debre Birhan University). Contact: besufekadbabibaw@gmail.com --- ## Features - **Exploratory Data Analysis (EDA)** with `pandas` & `scipy` - **Interactive visualizations** (monthly trends, origin pricing, warehouse volumes) - **Time-series forecasting** using `fbprophet` (5.2% error) - **Professional LaTeX report** with tables, figures, and citations - Actionable **recommendations** for producers, exporters, and policymakers --- ## Tech Stack - **Language**: Python 3.8+ - **Libraries**: - `pandas`, `numpy` – data processing - `scipy` – statistical analysis - `matplotlib`, `seaborn` – visualization - `fbprophet` – forecasting - **Environment**: Google Colab (Jupyter notebooks) - **Reporting**: LaTeX (`sn-jnl` template) - **Version Control**: Git & GitHub --- ## Repository Structure