Creating a reproduction package

So, you are ready to submit your paper. Congratulations! Now, it is also a good time to think about a reproduction package. The checklist below guides you through the process.

Create your package in (less than) 10 steps

1. Create a separate directory for your reproduction package. You can either clone/fork/download the template provided by this project or create a directory structure yourself. For example:

   linux:/data/user> git clone https://github.com/jdeplaa/open-data-template.git
   

2. If you performed your analysis in Jupyter Notebooks, you can include the relevant files in the notebooks directory. Please also include a short README file explaining how you need to run the Jupyter notebooks and which part of the analysis can be found in which notebook. If your notebooks are complete, you can skip step 3 and 4.

3. For each figure in your paper, copy the files/scripts that you used to create the figure to the figures directory. If you reserve one directory per figure, the relevant files are easier to find. In each figure directory, add a short README file or log to explain how to generate the figure using the files.

4. For each of your scientific results in the paper, copy the data end products and scripts to the data directory. Data end products can be spectra, images, tables, etc. If possible, also provide the scripts/software how you derived the scientific results from these end products (e.g. the fitting process). In the data directory, add a short README file or log how to derive the results using the scripts and data files.

5. Write a README.md file in the top level of the reproduction package. This will be the cover page for your package on Zenodo. In the file, list the software packages that you used and explain what can be found where in the package. The template README.md can guide you through this process.

6. Show your package to a colleague/supervisor. This is to check whether your package is clear enough for publication. Also check whether you did not include security sensitive data (usernames, passwords, computer names, etc.).

7. Tar or Zip the figures/data/notebooks directories. Do not include the README.md file in the top directory in the tar/zip package! The README.md needs to be outside the tar/zip package for Zenodo to display a preview of it.

8. Do a test upload of your package to the Zenodo sandbox. You can use your ORCID to login. Show the result to a colleague/supervisor to check. This upload will be removed after a couple of days.

9. If all the checks are OK, upload your package to the real Zenodo. You can use your ORCID to login. Do not forget to add your institute community to your upload, or any other community that you want to submit to. Be sure to let Zenodo create a DOI for you and do NOT use the DOI of the paper (What is a DOI).

10. Make clear in your paper that the data are available online. Some journals require you to add a data access statement to your paper and the link provides examples how to do that. If your journal does not require it, it is anyway a good idea to show in your publication where they can find the reproduction package.

Congratulations! You are done! :)

Useful links

• Introduction to MarkDown (formatting language of this file)
• More help creating repro packages.
• An example reproduction package.
• Zenodo sandbox.