A Practical Framework for Small Teams to Develop Sustainable Research Software
Bibliographic record
Abstract
Currently, a knowledge gap exists between research software and general software engineering. The scientific literature is full of ideas to close this gap, including documentation templates, code generation, continuous integration/deployment and formal methods. Although these ideas are promising, they often assume a large team that includes individuals who have the required expertise. Our proposed practical framework instead targets a small team of domain experts, with the only requirement being to find someone (either from the original team, or externally added) who is interested in deepening their software knowledge by volunteering for the developer role. Our framework, especially for the beginning stage of requirements elicitation, includes step-by-step guidance. The process begins with questions the developer asks the domain expert(s). These questions cover topics such as the expected inputs and outputs, the computational scale of the problem and special input cases with known solutions or trends. The methodology shows how to map the answers to these questions to the requirements, high-level design and verification documentation. Templates for all documentation, in markdown format, are provided in a GitHub template, along with the initial infrastructure for issue tracking and continuous deployment of the project's webpage. The proposed methodology incorporates four main pieces of advice: i) the notation and structure for documenting the theory should be selected to facilitate the transition to design and implementation; ii) continuous integration should be part of the project from the start; iii) the low-level design documentation should be done through structured comments in the code, like docstrings or doxygen; and, iv) the modular decomposition needs to consider the computational scale when balancing information hiding and performance.
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 | 0.082 | 0.065 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.005 | 0.003 |
| Science and technology studies | 0.008 | 0.015 |
| Scholarly communication | 0.013 | 0.017 |
| Open science | 0.009 | 0.020 |
| Research integrity | 0.010 | 0.006 |
| Insufficient payload (model declined to judge) | 0.021 | 0.012 |
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".