HaPy-Bug - Bug localization learning to rank algorithm results

# Replication Package for L2R Models Evaluation<br>This repository contains all necessary resources to replicate the experiments described in our evaluation of Learning-to-Rank (L2R) models applied to grayscale-based code change analysis.<br>## OverviewWe evaluate three L2R models for ranking modified lines of code based on their relevance:**Model 1**: Baseline L2R model.**Model 2**: Uses expert-annotated data.**Model 3**: Fully automated grayscale-based annotation model.<br>Model 3 automatically assigns labels to modified code lines based on grayscale heuristics, and its ranking results are generated similarly to Model 2.<br>### Results Summary<br>| Model | MAP | MRR | ACC@1 || Model 1 | 0.408 | 0.659 | 0.521 || Model 2 | 0.429 | 0.705 | 0.608 || Model 3 | 0.427 | 0.704 | 0.608 |<br>These results suggest that the automatic grayscale-based Model 3 achieves comparable performance to expert-labeled Model 2, at a much lower cost.<br><br>## How to Replicate<br>All files required to replicate the experiments (datasets, configuration, and scripts) are available at:<br>👉 https://doi.org/10.6084/m9.figshare.24448888