Publishing options
The main reasons to publish data is to share it with others and/or following a requirement from a publisher, funder or your own institution. The specific requirements are covered in the institutional policies and journal policies pages. While these requirements might differ in some details they are all based on the FAIR principles, we can help you making a decision on how to publish your data or code in a way that satisfy these principles. This will depend on the kind of data and the reasons you want to publish.
If you are unsure about exactly which data you should be publishing, the CLEX data policy and our guidelines should help you decide.
Whichever way you chose remember to report your record to CLEVER (see below).
Where should I publish my data?
There are at least three options and there is not a straight answer, it depends on what you are publishing and why.
CLEX Data Collection on NCI
This is the best option if you have data in NetCDF format and this data could be useful for other researchers. We will help you document your data and make it user friendly; it will be part of a climate data collection and so it will be easier to discover. NCI also has more storage capacity than other repositories and services which are designed around the NetCDF format.
Institutional repository
This is adequate if you have a small dataset and it is really specific to your study or only a subset/post processing of another dataset. While institutions offer some data curation, they usually will not check that the data is well described, consistent and user friendly, so you might get a DOI for your record but no added value. If your dataset is bigger than 50-100 GB, you might not be able to publish it with one of these repositories.
Zenodo, Figshare, Mendeley
These services are free, and you can create your own account, a record for your data and get a DOI fairly easily and quickly. You can also publish here different kind of materials. This can be useful if you want to publish some very specific data, for example code and data to produce a specific figure required to publish a paper.
However, there are no standards required or anyone checking on your metadata. This means that it is up to you to make sure your data is as FAIR as possible. This means Findable, which is harder when your record is not part of a discipline repository or collection or you haven't used keywords in an effective manner. And Accessible, Interoperable and Reusable, which means the data should have enough metadata, use discipline standards and be properly described.
Finally, as for institutional repositories, the data size is limited to 50 GB and you will not get any additional data services apart from HTTP download. If you decide to go this way, please make sure you document your data properly, we are happy to provide support and review your record. If you use Zenodo, you can easily add your record to our Data Collection (see below)
CLEX Data Collection
We have started a CLEX Data Collection in Zenodo to collect in one place all our data records, regardless of how they have been published. Having one place where all our data is listed means that it is easier for anyone to discover CLEX data outputs, both for external potential users and for our own researchers and students. If you have already published your data with your own institution and/or with one of the freely available services, like Zenodo itself, let us know and we will list your record in our collection. We will use your original metadata record, data access url and existing DOI (if available) as official source.
If you are confused feel free to ask us, we are always happy to provide advice and support.
Where should I publish my code?
Like data, code represents part of your work and funders are starting to look at all research products not only papers when reviewing grant applications, also some journals require you to publish your code alongside the data. Putting your code on GitHub or another version control service helps to keep track of the code, expose it to others and manage potential issues and enhancements. However, GitHub is not ideal if you want to pinpoint the code you used for a paper or to create some data.
We started a Zenodo community for a CLEX Code Collection (CCC) in 2020. Zenodo is a platform that will mint a DOI for your code and integrates well with GitHub. Initially we published some of our own codes and code used to produce papers as required by journal editors. We are now looking into broadening this and actively seek contributions of code, notebooks etc that might be useful to others. Zenodo has given our codes much more visibility than GitHub and some of the codes have lots of views and downloads.
Publishing options diagram
How to publish
We are not providing here specific advice if you are publishing with an institutional repository or other non-specific repositories. However, the steps involved in both preparing the files and filling in the metadata form will be very similar to what we cover in "Publishing data in Zenodo". So, you can refer to that wiki page and the step by step guide linked into it for useful advice.
- Publishing with NCI
- Publishing with your institution
- Publishing code in Zenodo CLEX Code Collection
- Publishing data in Zenodo CLEX Data Collection
Reporting to CLEVER
Whichever way you decide to publish your data, as part of the NCI collection or with a repository provided by your institution, remember to add your published record to CLEVER the CLEX reporting hub, in the "Publications and Datasets" section.
You will need to record only the main information: author, title, DOI and citation. It will only take a couple of minutes. Published datasets are part of the Centre KPIs and something we have to report to our funders.