A new feature called the Asset Repo has now been added to DataLabs which is fully operational. The goal of this is to curate a central set of datasets/code/links which are useful to the community. In many cases when a dataset is required for analysis file-level access is required to interrogate the data, this means that users have to upload the data to within the DataLabs infrastructure. This is expected however in many cases this causes the same dataset to be uploaded multiple times across users and projects. The creation of the Asset Repo has the advantage of being an area where users can access content while also preventing multiple copies of datasets being brought into DataLabs, as well as providing the community with a growing array of potentially useful datasets.


The main points can be found below. Further guidance on how the Asset Repo works and can be used within notebooks/sites can be found on the following page https://datalab-docs.datalabs.ceh.ac.uk/howto/use-the-asset-repo/

  • A new tab called Asset Repo is now accessible in the Top Bar of DataLabs which contains a list of curated datasets.
  • Datasets can either be made public to the entire community or restricted to one or a number of projects depending on the need.
  • Attached datasets once attached can be accessed directly within either Notebooks or Sites in a read-only fashion.

In order to get things rolling we have added the following datasets below which are now available for all users. 


We encourage users to let us know if there is a dataset they would like to see included which can be added (this currently applies only to publicly licensed datasets), or if there is a dataset that has been produced within DataLabs and is of value to other projects.

IMPORTANT NOTE: In many cases the data held within DataLabs will be a copy of a dataset held within EIDC or another source repository. If this is the case a DOI will be listed on the Asset Repo entry which can be used to see the raw data page as well as information on how to cite the data if necessary.

  • No labels