More detailed catalog information - Githubissues

pangeo-forge / user-stories

User stories to guide PF development

1 stars 0 forks source link

More detailed catalog information #12

Open rabernat opened 2 years ago

rabernat commented 2 years ago

User Profile

As a data library user

User Action

I want to be able to see detailed information (e.g. variables, dimensions, attributes) about the datasets in the catalog

User Goal

So that I can decide whether I want to use a particular dataset.

Acceptance Criteria

[ ] User browses to https://pangeo-forge.org/catalog an can click on a dataset (related to #1)
[ ] Information is shown similar to the Xarray repr, or the legacy Pangeo catalog (e.g. https://catalog.pangeo.io/browse/master/ocean/SOSE/)
[ ] Code samples show how to open the data in python.

Linked Issues

1

rabernat commented 2 years ago

I think it is useful to think of the storage target location as a "data lake". So the key questions here are

What services can we expect from the data lake? Currently we just assume that these are Zarr datasets in a bucket. But could the DL expose the metadata about the datasets via an API?
How should Pangeo Forge be identifying datasets in the DL? Currently we just pick a path (see ADR-3. However, we could use a different identifier system.

abkfenris commented 2 years ago

xpublish could be useful here as it exposes both the xarray repr for datasets along with a NCO-JSON representation.