HR Analytics System

Data cleaning + metrics calculation system built for Wits PGDip Programming for Data Science.

What This Project Does

Component	What It Does
cleaner.py	Removes null salaries, standardizes department names, validates performance ratings (0-5), fixes date formats, validates hire dates (2015-2025) with leap year logic
metrics.py	Calculates retention rate, turnover by department, average age, average salary by age range, and performance by training hours
sql_cleaning_queries.sql	Same cleaning logic written in SQL – adapted for equipment maintenance data (MTBF, uptime, failure rates)

Skills Demonstrated

• Python data cleaning (null handling, validation, date logic)
• Business metric calculation from raw data
• SQL equivalents for ERP/maintenance databases
• Edge case handling (leap years, empty datasets, out-of-range values)
• Testing with Pytest

Industrial Transferability

This HR analytics logic applies directly to equipment maintenance data:

HR Concept	Maintenance Equivalent
Retention rate	Equipment uptime percentage
Turnover rate	Failure rate by component
Average salary by age	Average MTBF by equipment class
Performance by training	Performance by maintenance hours

How to Run

# Clone the repository
git clone https://github.com/RebeccaDataSci/hr-analytics-system.git

# Generate fake data
cd hr-analytics-system/data
python generate_fake_hr_data.py

# Run cleaning and metrics
cd ../src
python cleaner.py
python metrics.py

# Run tests
pytest tests/

## Project Context

- **Course:** PGDip Data Science, University of the Witwatersrand
- **Module:** Programming for Data Science Introduction
- **Date:** March/April 2026
- **Group:** Group 11 (my contributions: data cleaning module, metrics functions, report writing)

## Author

**Rebecca** – PGDip Data Science @Wits | Former electrician, Anglo American smelter

[GitHub Profile](https://github.com/RebeccaDataSci)