Add documentation about catalog configurations

[klieberman]

It may be helpful to add documentation about how to customize catalog configurations. Since the DataCatalog class already provides a json_uri parameter to allow custom configuration files, I think providing example configuration files in the repo and/or documentation about the required format of configuration files would be useful. For example, I downloaded the data for one sensor to our system (for more efficient training) and then needed to figure out where/what the current configuration file was (catalog.json in the google cloud bucket under nnja-ai/data/v1-preview), add it to my local repo, and make some guesses about what values to change. While this was fairly straightforward, I imagine this is a common use case and documentation could make this process simpler.

[linear]

DAT-128 Add documentation about catalog configurations

[hansmohrmann]

Hi @klieberman! So if I understand correctly, your use case is "I want to use this SDK for interfacing with a local copy of the data" and right now there's no obvious way to adapt the various config files to support that (though you've managed to make it work). Is that right?

[klieberman]

Yes, exactly. It's just a suggestion to make the code more usable, feel free to ignore. Thanks for building a great repo!

[hansmohrmann]

It's a good idea to keep in mind. I hadn't considered that use case but can definitely see the value. We have to make some changes to the catalog spec to better support multi-cloud, so I'll try to make a good example for working with the data locally as well!

[hansmohrmann]

Long time coming, but this PR and the resulting notebook should address this!