Making your project citable — GitHub without the command line documentation (original) (raw)

Discussion: is depositing your data/code on GitHub enough?

There are many services where you can share or archive your code and data: See for instance ourlesson on reproducible research.

In this present lesson we will discuss one of the many options to get a **digital object identifier (DOI)**for your dataset or code:Zenodo, A general-purpose open access repository created by OpenAIRE and CERN.Zenodo has nice integration with GitHub, and allows researchers to upload files up to 50 GB.

We will exercise in the Zenodo sandbox

We will practice on https://sandbox.zenodo.org/and not on the “real” https://zenodo.org/to make sure we do not create “real” DOIs which we cannot remove.

The sandbox service is useful to calibrate your setup until you are happy with the result and then you can go for the real service. Once a dataset is uploaded to the “real” service, it cannot be easily removed or modified again (and this is good, otherwise DOIs would not make much sense).


Step 1: Prepare an example repository

Through web:

Or using GitHub Desktop:

Alternatively we can also practice this with one of the repositories we created earlier in this lesson.

Here I just created a new one:

../_images/example-repo.png


Step 2: Activate the repository on Zenodo (sandbox)

../_images/zenodo.png

../_images/switch.png

../_images/sync-now.png


Step 3: Create a “release” and get a DOI

../_images/releases.png

../_images/publish-release.png

../_images/zenodo-published.png


Step 4: Add a DOI badge to your repository

This is bonus but for visitors of your GiHub repository it can be nice to find a badge in your README that informs them about and links to the preserved dataset/code on Zenodo.

../_images/zenodo-badge.png

../_images/github-badge.png