Bibliographic record
Abstract
A study on reproducibility smells This study, first, identifies such programming practices that we refer to as reproducibility smells by conducting a comprehensive multi-vocal review and propose a first-ever validated catalog of reproducibility smells for IaC scripts. We implement a tool viz. REDUSE to identify reproducibility smells in Ansible scripts. Furthermore, we carry out an empirical study to reveal the proliferation of reproducibility smells in open-source projects and explore correlation and co-occurrence relationships among them.We elaborate on the contents of the individual folder of this repository below. Multi-vocal literature review(mlr) This folder contains the following files. grey-literature-source.xlsx: This Excel file contains all the information about the grey literature resources. It has following sheets. Links: contains information about the type of the documents, content of the document, what IaC language or concept it is related to and the corresponding link to the document. Reviewer-1-QA: contains the marking assessment of each document from the reviewer-1 for the grey literature. Reviewer-2-QA: contains the marking assessment of each document from the reviewer-2 for the grey literature. Grey Literature Sources: contains the source numbers, best practice/ bad practice that each of the literature are mentioning and a summary. Search Queries: shows that how many documents we gain for each search query, how many filtered with the exclusion/inclusion criteria and number of final documents. smell-reference: shows a list of references that each of the smells are extracted from the them. Quality Assessment final: contains the final and merged marking assessment of each document from the both reviewers for the grey literature. smell-examples.txt: contains the example code for each of the reproducibility smell discussed in the paper. smell-descriptions.pdf: contains the detailed description for each of the reproducibility smell discussed in the paper. REDUSE - a reproducibility smells detection tool (reduse) The tool is designed to detect reproducibility smells in Ansible scripts.You can provide your ansible script in .yml format and get output of the tool as a csv file containing task name, task number, smell name, reason of having the smell on the task. Build/Configure This tool requires Python 3.8+ Install the packages from reduse\src\requirements.txt Run the tool Run the detector.py file in the reduse\src folder with the path to your desired Ansible yaml file. python detector.py '/path/to/file/folder' Or, add the path to the desired ansible files or repositories to the script and then execute /bin/bash run_detector.sh Manual validation (reduse\manual-validation) This folder contains all the scripts i.e., the subject systems and detection smells by the tool as well as by the evaluators. scripts: Selected scripts for evaluation results: Detected smells by human evaluators and the tool. Each subfolder contains two kinds of files. Files ending with manual are produced by human evaluators whereas files ending with tool are generated by the tool. Empirical study (empirical-study) Analysis scripts (analysis-scripts) This folder contains all the scripts that are used to calculate the metrics required for the empirical study, such as the smell frequency, correlation, and confusion matrix. Resources (resources) extracted-repos: this folder contains the list of the open-source repositories from nine categories hosted by the Ansible Galaxy platform. extraction-scripts: scripts to check the criteria and clone the repositories from the Ansbile Galaxy platform. repos: contains Ansible scripts for all the selected repositories used for the empirical studies. Qualitative analysis (qualitative-analysis) Ansible-galaxy-issues.pdf: contains links to the repositories, issues, and the justification for the issue related to reproducibility smell from the qualitative analysis section. Qualitative-analysis.xlsx: contains information such as repository name, issue number, root cause, smell description, and link of the issue for all the issues considered for this analysis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.052 | 0.245 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.003 | 0.005 |
| Bibliometrics | 0.007 | 0.006 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.009 | 0.010 |
| Open science | 0.005 | 0.007 |
| Research integrity | 0.002 | 0.005 |
| Insufficient payload (model declined to judge) | 0.317 | 0.223 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".