This project is Task 2 of the CodeAlpha Data Science Internship. It performs a full-scale analysis of unemployment data across India from May 2019 to October 2020, with a sharp focus on how the Covid-19 pandemic and the March 2020 national lockdown affected employment across all states, zones, and area types (rural/urban).

The analysis answers real policy questions: Which states were hit hardest? Did cities suffer more than villages? What does the recovery look like?

• ✅ Clean, preprocess, and merge two complementary unemployment datasets
• ✅ Perform thorough EDA with monthly, regional, and area-wise visualizations
• ✅ Quantify and visualize the Covid-19 impact on unemployment rates
• ✅ Analyze zone-wise and state-wise trends using heatmaps and comparisons
• ✅ Compare Rural vs Urban unemployment before and after lockdown
• ✅ Build a time series forecast model to project recovery trajectory
• ✅ Derive actionable policy insights from the data

Features used across both datasets:

💡 Exact numbers generated when you run the notebook — results printed in the final summary cell.

1. Covid-19 caused the sharpest unemployment spike in the dataset — The March 2020 lockdown triggered a dramatic and near-instantaneous rise in unemployment across all states.

2. Urban unemployment was consistently higher than rural — India's urban informal workforce (daily-wage, gig workers) bore a disproportionate burden.

3. Labour Force Participation dropped sharply — Many workers, especially women, exited the workforce entirely during the lockdown rather than remaining "unemployed."

4. Recovery was underway by June 2020 — The polynomial regression forecast confirms a declining trend from the peak, though rates remained above pre-Covid levels.

5. State-level variation is significant — Some states maintained relatively low rates throughout, while others hit extreme highs — highlighting the need for targeted interventions.

1. Expand rural job guarantee schemes (like MGNREGS) automatically during national crises
2. Portable benefits for urban informal workers — delink social security from employers
3. State-specific interventions for consistently high-unemployment states
4. Women's workforce re-entry programs to recover dropped Labour Force Participation
5. Real-time monthly unemployment dashboards for faster policy responses

CodeAlpha_UnemploymentAnalysis/
│
├── 📓 unemployment_analysis.ipynb       ← Main Jupyter Notebook
├── 📄 README.md                         ← This file
├── 📋 requirements.txt                  ← Python dependencies
├── 📂 Unemployment_in_India.csv         ← Dataset 1
├── 📂 Unemployment_Rate_upto_11_2020.csv ← Dataset 2
│
└── 📊 Generated Plots/
    ├── national_trend.png
    ├── distribution_analysis.png
    ├── seasonal_patterns.png
    ├── covid_impact.png
    ├── zone_analysis.png
    ├── state_heatmap.png
    ├── top_bottom_states.png
    ├── rural_urban_analysis.png
    └── forecast.png

Important: Upload both CSV files when prompted in Colab (or place them in the same folder as the notebook).

# 1. Clone the repository
git clone https://github.com/MOHAMMED-ABUZAR317/CodeAlpha_UnemploymentAnalysis.git
cd CodeAlpha_UnemploymentAnalysis

# 2. Install dependencies
pip install -r requirements.txt

# 3. Launch notebook
jupyter notebook unemployment_analysis.ipynb

pandas>=2.0.0
numpy>=1.24.0
scikit-learn>=1.3.0
matplotlib>=3.7.0
seaborn>=0.12.0
jupyter>=1.0.0

pip install -r requirements.txt

• How to clean and merge multiple real-world datasets with different structures
• Visualizing time series data with event markers (lockdown dates, shaded periods)
• Quantifying the economic impact of a real-world event (Covid-19) with data
• Building polynomial regression for non-linear time series forecasting
• Translating data insights into actionable policy recommendations