Dataset for Code Classification, Transformation, and Recommendation Tasks

This repository contains curated datasets designed for research on programming style, code quality, and sociolinguistic factors. The data was collected from two major programming platforms: Codeforces and GitHub. Each dataset is stored in a CSV file and is annotated with various metadata attributes such as gender, region, expertise, and code compliance.

These datasets are suitable for tasks such as:
- Code quality prediction using LLMs
- Demographic and regional analysis of coding styles
- Program transformation (e.g., converting non-compliant to compliant code)
- Fairness-aware ML in software engineering

Each file is ready for ingestion into machine learning pipelines or statistical analysis tools. The code snippets are executable, and maintainability scores have been validated using Halstead metrics and parsing tools.