Open MathewBiddle opened 2 years ago
I think we would probably want to wait until after the GSOC refactoring to work on this, but maybe it's a good code sprint project.
I think we would probably want to wait until after the GSOC refactoring to work on this, but maybe it's a good code sprint project.
Indeed. However, we can start discussion this high level API now. For example, would something like this make sense?
from ioos_erddap import ERDDAP
ds = ERDDAP(ra="secoora")
ds ← list all dataset_ids
ds["dataset_id"] ← return a xarray object for that dataset using the OPeNDAP response to take advantage of the lazy loading.
and
ds = ERDDAP(ra="secoora", **constraints)
ds ← list all dataset_ids constrained in space/time/query string
ds["dataset_id"] ← return an in memory/fsspec xarray object for the constrained dataset
@MathewBiddle do you believe it is worth adding this as a topic for the code sprint this year?
Yes!
may i know what the code sprint is all about? and is this issue still open...?
Hi @Harish-2003, you may find more information on the code sprint at https://ioos.github.io/ioos-code-sprint/
The short version is: The IOOS code sprint is where some of us sit together to try to solve a collection of selected issues and/or add features to our Software stack.
Could we use the existing code base of ioos/colocate as a starting point?
Could we use the existing code base of ioos/colocate as a starting point?
This would be lower level. The goal would be for colocate to be a front-end for it.
Similar to gliderpy and various other wrapper packages, we should think about an IOOS wrapper package for high level queries against the various IOOS ERDDAP endpoints. Maybe provide some high level utilities too?
A subset of RA endpoints are documented here [1]. However, it would probably be better to use the awesome-erddap json instead [2]. Or, identify another curated list.
[1] https://github.com/ioos/ioos-asset-inventory/blob/main/utils/ra_erddaps.json
[2] https://github.com/IrishMarineInstitute/awesome-erddap/blob/master/erddaps.json