Reproducibility Checklist: Data

[LilacHo]

Data (FAIR, CARE)

Extra resources:

• OHW23 Joseph Gum's presentation
• Software Best Practices, Valentina Staneva's presentation

• Are the data publicly available? If not all, can a summary of them be made publicly available?

  
  - If space is an issue you can subsample the data for demonstration purposes, or store at least an intermediate dataset.
  
• Are they in a format easily accessible by open source software libraries or following a community convention?

  
  - Storing data in library-specific formats such as .npy, .mat, or .RData can prevent users from accessing the data. Switch to library-agnostic formats such as .csv,.nc, etc.
  - Check if there is a community convention for storing the data within your field? Contribute feedback.
  
• Are the data stored with the corresponding metadata?

  
  - Ensure the versions of the data and metadata align.
  
• Do they have a license that permits broad use?

  
  - Learn more about data licenses
  - You can state people using your data cite your project
  - If the data was obtained from a partner, what is the data sharing agreement?
  
• Are they permanent, or do they have versions?

  
  - Do you specify the version or the time stamp when the data were retrieved?
  - Note DOIs can have versions
  
• Do they have a DOI?

  
  - You can create DOI for your own dataset using Zenodo, Dryad
  
• For how long can they be stored at their current location?

  
  - If you used a temporary location, but retrieved from a public repository, provide the steps on how you retrieved the data.
  

[LilacHo]

• Are the data publicly available? If not all, can a summary of them be made publicly available?

  
  - The data is available on Google Cloud Storage (GCS). It is not publicly available but it could be accessed by people who has the correct access.
  
• Are they in a format easily accessible by open source software libraries or following a community convention?

  
  - The data is in the .zarr format, which is easily accessible.