HR Attrition Dashboard

This project is an interactive HR Attrition Dashboard built using Jupyter Notebook, ipywidgets, matplotlib, and seaborn. It allows users to explore key HR metrics such as attrition, demographics, job satisfaction, and employee engagement.

The dashboard behaves like a web application, letting users filter employees dynamically by:

• Department
• Gender
• Age Range

All visualisations update instantly as filters change.

---

🎯 Objectives

Analysing attrition helps companies understand:

• Predictive factors of attrition: Identify which characteristics (e.g., age, work-life balance, job role) influence employee turnover.

• Areas for intervention: Spot potential improvements in HR practices such as workload balance, career progression, or employee well-being.

• Employee engagement and satisfaction: Understand correlations between job satisfaction, environment satisfaction, and attrition risk.

---

🗃️ Dataset

File: data/hr_analytics.csv

The dataset used in this project is a HR Employee Attrition dataset, originally found on Kaggle (exact source link not available anymore).

It includes 1,480 employee records with fields such as:

• Demographics: Age, Gender, Marital Status
• Work environment: Job Role, Department, Business Travel
• Satisfaction metrics: Job Satisfaction, Work-Life Balance
• Performance attributes: Performance Rating, Years at Company
• Attrition status: Whether the employee left the company

---

🚀 Key Features

1. Real-time interactivity: Built with ipywidgets, the dashboard reacts instantly to filter inputs, no rerunning cells required.

2. Multiple visualisations: Donut charts, (horizontal/vertical/clustered) bar charts, violin plot

3. Clean, user-friendly UI: Using VBox, HBox, and structured layout design, the filters and charts are arranged into a dashboard-like interface that feels similar to BI tools (Power BI/Tableau).

4. Voilá-compatible: The notebook is fully structured for Voilá deployment, turning it into a standalone web application without exposing any notebook code.

---

📁 Project Structure

hr_attrition_dashboard/
│
├── data/
│     └── hr_analytics.csv            # Dataset
│
├── media/
│     └── dashboard_demo.gif          # A short demo of the dashboard
│
├── .gitignore                        # Files/folders excluded from Git
│
├── LICENSE                           # MIT License
│
├── README.md                         # Project overview and instructions
│
├── environment.yml                   # Conda environment for local setup
│
└── hr_attrition_dashboard.ipynb      # Main dashboard notebook

---

🔧 Setup & Execution

This project can be run in several ways depending on preference, from local development to cloud execution. The options below are arranged from most flexible to most convenient.

Option 1: Run Locally (Recommended for Development)

1. Clone the repository

   git clone https://github.com/ChokZB/hr_attrition_dashboard.git
   cd hr_attrition_dashboard

2. Install dependencies

   conda env create -f environment.yml
   conda activate hr-attrition-env

3. Run the notebook

   jupyter notebook hr_attrition_dashboard.ipynb

   The dashboard interface will appear inside the notebook after running all cells.

---

Option 2: Run Locally with Voilá (View as a Clean Web App)

Voilá turns the notebook into a standalone web application with no visible code.

1. Install Voilá

   pip install voila

2. Launch the dashboard

   voila hr_attrition_dashboard.ipynb

   A browser window will open displaying the dashboard as a full interactive app.

---

Option 3: Run in Google Colab (No Installation Needed)

Open the notebook directly in Colab:

---

💡 Use Case Highlights

This project demonstrates:

• Building interactive dashboards entirely in Python
• Web-app-style interfaces with ipywidgets + Voilá
• Data preprocessing, visualisation, and UI engineering
• Real-world HR domain analytics
• Visual storytelling with multiple chart types
• An alternative to BI tools using only code

---

👨‍🏫 Dashboard Demo

---

🧑‍💻 Author

Chok Zu Bing

GitHub: @ChokZB

---

🪪 Licence

This project is released under the MIT License.