Dataset for Empirical Study: Code smells and Refactorings

The dataset consists of the following archives :

smell_ref_map.zip :

*.json - Smell instances with mapped refactorings + unmapped refactorings for each repository's version history

*.chain.json - Chained smell instance IDs with respective *.json after postprocessing

*.stats.json - Smell instances and refactorings statistics.

corpus.csv - Chained smell instances in a dataframe-friendly format for easy analysis

manual_analysis.zip - Top 5 smell-refactorings samples analyzed by a human and Large Language Model (Total of 802 samples)

manual_analysis_for_unmapped.zip - Smells without any removal refactorings mapped (Total of 270 samples)

* = repository name in format "repository_name@username" hosted at GitHub