🧹 Data Cleaning & Exploratory Data Analysis (EDA) on Global Layoffs Dataset

---

📑 Table of Contents

• Project Overview
• Dataset Information
• Data Cleaning Process
• Exploratory Data Analysis (EDA)
• Key Insights
• Usage
• Example Queries
• Repository Structure
• Visualizations
• Future Improvements
• Contributing
• License
• Author & Contact

---

📘 Project Overview

This project focuses on data cleaning and exploratory data analysis (EDA) of a layoffs dataset covering global companies between March 2020 – March 2023. The raw Excel dataset was imported into MySQL, where all cleaning, transformation, and analysis were executed using SQL queries.

---

🗂️ Dataset Information

• Raw file: raw_layoffs.xlsx
• Imported table: layoffs_staging2
• Period covered: 2020-03-11 → 2023-03-06

Columns:

Column	Description
company	Name of the company
country	Country of operation
date	Layoff announcement date
sector	Industry or business sector
total_laid_off	Employees laid off
percentage_laid_off	Workforce percentage affected
location	Company location
stage	Company stage (Startup, Series C, Public, etc.)
funds_raised_millions	Total funds raised in USD millions

---

🧽 Data Cleaning Process

Key cleaning operations performed:

1. Removed null, duplicate, and inconsistent records.
2. Standardized date format (YYYY-MM-DD).
3. Trimmed whitespaces and fixed casing for text columns.
4. Replaced blank percentage_laid_off with NULL.
5. Converted funds_raised_millions to numeric data type.
6. Created a cleaned table for analysis.

---

🔍 Exploratory Data Analysis (EDA)

Executed directly in MySQL using GROUP BY, JOIN, WINDOW FUNCTIONS, and CTEs to explore:

• Monthly and yearly layoff trends
• Top companies with highest layoffs
• Layoffs by country, sector, and stage
• Correlation between funds raised and layoffs
• Average layoff percentage per sector

---

📈 Key Insights

• Peak layoffs occurred in mid-2020 and early 2023.
• Tech and consumer-services sectors were hit hardest.
• Startups in late-funding stages had higher layoff rates.
• Some highly funded firms still made massive cuts — funding ≠ stability.

---

💻 Usage

🔧 Requirements

• MySQL 8.0+
• Any SQL client (MySQL Workbench / CLI)
• Raw dataset: raw_layoffs.xlsx

▶️ Steps to Run

1. Clone this repository

   git clone https://github.com/Aadityavarier/Data-Cleaning-and-EDA.git

2. Convert Excel file to CSV for MySQL import (if needed)

   • Open raw_layoffs.xlsx and save/export as raw_layoffs.csv.

3. Import the raw dataset into MySQL

   LOAD DATA INFILE 'path_to/raw_layoffs.csv'
   INTO TABLE layoffs_staging2
   FIELDS TERMINATED BY ',' 
   IGNORE 1 ROWS;

4. Run the SQL script

   mysql -u root -p < "data_cleaning_and_eda_project.sql"

5. Explore outputs and insights using SELECT statements.

---

🧾 Example Queries

-- 1. Top 5 companies with highest layoffs
SELECT company, SUM(total_laid_off) AS total
FROM layoffs_staging2
GROUP BY company
ORDER BY total DESC
LIMIT 5;

-- 2. Monthly layoffs trend
SELECT DATE_FORMAT(date, '%Y-%m') AS month,
       SUM(total_laid_off) AS total
FROM layoffs_staging2
GROUP BY month
ORDER BY month;

---

🗂️ Repository Structure

📁 data-cleaning-and-eda-project/
│
├── 📄 raw_layoffs.xlsx               # Original dataset (Excel)
├── 📄 data_cleaning_and_eda_project.sql  # MySQL cleaning + EDA queries
├── 📄 LICENSE                        # MIT License
└── 📄 README.md                      # Project documentation

---

📊 Visualizations

**Below: Chart generated from query outputs. **

To see more go to (Visuals)

---

🤝 Contributing

Contributions and suggestions are welcome!

1. Fork the repo.
2. Create a feature branch.
3. Submit a pull request.

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.

---

👤 Author & Contact

Aaditya V
B.E. in AI & Data Science (2nd Year) — Mumbai University

• Email: aadityav1703@gmail.com
• LinkedIn: aadityavarier
• GitHub: Aadityavarier