Docs menu
- Home
- Docs
- Requirements
- Access & preservation
Access & preservation: lasting, citable materials
Access describes how others can obtain your materials; preservation keeps them available after the paper. Both sit beside the level, with license, citation and journal policy.
Last updated
Access conditions from the latest check
- Some data available with registration
- Some data available by application
- Requires commercial software
Archiving not assessed · Separate from the readiness level.
A repository can be renamed, rewritten or deleted after publication. Journals ask for durable copies: PLOS Computational Biology recommends depositing code in a repository that issues DOIs. RepoReady reports the access conditions it finds; it does not verify archives yet. Journal policy and good-practice findings are Warnings: they never change the level.
In this guide
Access conditions
- What counts
- Each dataset and tool says how it can be obtained: openly, after registration, through a data access committee, by purchase or under license.
- Common gaps
- “Available on request” with no route, or private data with no word on who may use them.
- How to fix
- State the route per item in the README. Public hosting, controlled access and private access are reported separately.
A lasting archive
- What counts
- Code and data sit in an archive that issues persistent identifiers, such as Zenodo, figshare, Software Heritage or a domain repository.
- Common gaps
- A GitHub repository alone, or a DOI that points to the newest version instead of the one in the paper.
- How to fix
- Connect the repository to Zenodo, so each GitHub release gets its own version DOI, and cite that version DOI.
Journal policy
- What counts
- What journals ask for beyond the level: a software license, a citation file that names the paper, source data for every figure and a code availability statement.
- Common gaps
- No license or only CC-BY on code, a dependency license that forbids reuse, no citation file, or empty, incomplete or duplicated source data.
- How to fix
- Deposit one source data file per figure, use our statement template, and add a license (without one, nobody may reuse the code) and a citation file:
cff-version: 1.2.0
message: "If you use this code, please cite the paper below."
title: "Code for: <paper title>"
version: "v1.0-paper"
doi: "<version DOI from Zenodo>"
license: MIT
authors:
- family-names: "<Family>"
given-names: "<Given>"
preferred-citation:
type: article
title: "<paper title>"
authors:
- family-names: "<Family>"
given-names: "<Given>"
doi: "<paper DOI>"Good practice
- What counts
- Advice beyond the levels, every finding a Warning: a tidy repository, tests, notebooks run in order, stated hardware and tools, and a paper that agrees with itself.
- Common gaps
- Junk or duplicate files, no tests, notebooks saved out of order, unstated GPU needs, commercial-only steps, or code older than the paper.
- How to fix
- Each Warning names the file or passage. The guides show fixes for a tidy repository, tests and notebooks and a consistent paper.
Related
See where your project stands
Run the analysis on your paper or repository. Every finding names its area and links back here.