Write a Blog >>
ICSE 2023
Sun 14 - Sat 20 May 2023 Melbourne, Australia

The paper, “PYEVOLVE: Automating Frequent Code Changes in Python ML Systems,” accepted for the ICSE-2023 technical track, introduces a novel technique for automating repetitive code changes in Python-based software systems. The proposed tool, along with the accompanying evaluation dataset, is easily accessible to the public through the tool’s repository on GitHub (https://github.com/maldil/PyEvolve), Artifact website (http://pythoninfer.github.io) as well as through Zenodo. The provided documentation and instructions in the tool’s repository ensure that the tool is easy to verify, use, and reuse. Therefore, we would like to claim the badges of “Artifact Available” and “Artifact Reusable” as a testament to the tool’s accessibility and usability. We believe that reviewers who are expected to be proficient in CLI command execution, managing VirtualBox VM images, and working with Mac OS will be able to evaluate and use the tool effectively. The image file is 19GB in size, so we expect the reviewers to have a good internet connection to download the image and at least 40-50 GB free space on their machine to load the image into VirtualBox. In order to speed up the review process, we kindly request that the reviewers begin downloading the image using one of the links provided at https://github.com/maldil/ICSE2023_PyEvolve_Artifacts/blob/master/README.md#1-tool—pyevolve (check step 1.2) before proceeding with the detailed steps, as this may take some time.

VirtualBox images are known to have compatibility issues when loaded on different operating systems. In addition, VirtualBox images may not be compatible with Mac computers that use different types of chips, such as Intel chips or M1/M2 chips. Our VirtualBox image was created on a Mac with an Intel chip, and for this reason, we expect the reviewers to use it on a Mac computer with an Intel chip for the best experience and smooth evaluation of the tool. All the required steps and details for artifact evaluation can be found in this link https://github.com/maldil/ICSE2023_PyEvolve_Artifacts.