🧹 SQL Data Cleaning on Global Layoffs

This project focuses on cleaning a dataset of company layoffs to prepare it for further Exploratory Data Analysis (EDA).
The data cleaning was performed entirely in SQL (MySQL).

---

📌 Project Overview

The raw layoffs.csv dataset contains inconsistencies such as:

• Duplicate rows
• Extra spaces in text fields
• Inconsistent industry and country naming
• Invalid or empty date formats
• Null values in certain columns

Through systematic SQL transformations, these issues were resolved and a clean dataset was produced:
👉 cleaned_data/clean_data_layoffs.csv

---

Dataset

• Source: Public dataset of global layoffs (CSV imported into SQL database).
• Columns include:
  • company – Name of the company
  • location – City or region of the company
  • industry – Industry sector
  • total_laid_off – Number of employees laid off
  • percentage_laid_off – Layoff percentage
  • date – Date of layoff
  • date_added – Date when record was added
  • stage – Company funding stage
  • country – Country of the company
  • funds_raised – Funds raised by the company
  • source - information source

---

⚙️ Steps in Data Cleaning

1. Back up the Original Data

• Created a backup table layoffs_duplicate to preserve the original dataset before manipulation.

2. Remove Duplicates

• Identified duplicates using ROW_NUMBER() in a CTE partitioned by all key columns.
• Removed duplicates by keeping only the first occurrence.

3. Standardize Data

• Trimmed whitespace from company names.
• Standardized country names (e.g., united arab emirates → UAE).
• Converted date columns (date, date_added) to SQL DATE type.
• Converted numeric columns:
  • total_laid_off → BIGINT
  • percentage_laid_off → DECIMAL(5,1)
  • funds_raised → BIGINT

4. Handle Null Values

• Converted empty strings to NULL.
• Populated missing country values using corresponding location.
• Checked and handled nulls in industry column.
• Deleted rows where both total_laid_off and percentage_laid_off were null.

5. Drop Unnecessary Columns

• Dropped temporary columns such as rn used for deduplication.

6. Final Output

• Exported the cleaned dataset into cleaned_data/clean_data_layoffs.csv.

---

📂 Repository Structure

SQL-Data-Cleaning-on-Global-Layoffs-Dataset/
│── README.md
│── Data_Cleaning_Project_using_layoffs_data.sql
│
├── data/
│   ├── layoffs.csv # Raw dataset
│   └── cleaned_data/
│       └── clean_data_layoffs.csv # Final cleaned dataset
│
├── images/
│   ├── before_cleaning.png
│   └── after_cleaning.png

---

🖼️ Before vs After Cleaning

Before Cleaning	After Cleaning

---

🚀 How to Use

1. Clone this repository:

   git clone https://github.com/SAHFEERULWASIHF/SQL-Data-Cleaning-on-Global-Layoffs-Dataset.git
   cd Data-Cleaning-Layoffs-Project

2. Load the SQL script:

   source Data_Cleaning_Project_using_layoffs_data.sql;

3. Access the cleaned data from:

   data/cleaned_data/clean_data_layoffs.csv

🛠️ Tools & Technologies

• SQL (MySQL) – For data cleaning, transformation, and manipulation.
• SQL Queries Used:
  • CREATE TABLE / INSERT INTO
  • ROW_NUMBER() with OVER(PARTITION BY ...) for deduplication
  • UPDATE and TRIM() for standardization
  • ALTER TABLE for modifying data types
  • DELETE to remove incomplete records

Final Outcome

• Cleaned dataset: clean_data_layoffs.csv
• Row count after cleaning: 3,458 rows
• Dataset ready for Exploratory Data Analysis (EDA) and visualization.

📊 Future Work

Perform EDA on the cleaned dataset.

Create visual dashboards (Power BI / Tableau).

Apply predictive analytics on layoff trends.

✨ Author

F SAHFEERUL WASIHF

• 🎓 Graduate (Fresher)
• 📊 Interested in Data Analytics, Software Development, and Data Cleaning
• 🛠️ Skills: SQL, Python, Data Visualization, Data Wrangling
• 🔗 GitHub | LinkedIn | Portfolio
• 📧 Contact: wasihfwork@gmail.com