Closed benbovy closed 4 years ago
This is ready ;-)
This is ready ;-)
Looks good.
One thing: With
requests.get(server_url + '/api/v1.0/datasets/demo/fieldnames').json()
lists all keys (including coords) as fieldnames, because https://github.com/ESM-VFC/esm-vfc-api-demo/blob/1355452d3f60fd5427349b9dbc786282cec7baac/app/app.py#L35 uses the Zarr keys for variable name lookup. I'd change to
fieldnames = list(xr.open_zarr(demo_dataset_http_map, consolidated=True).data_vars.keys())
Agreed, it's better to return data variables only.
I made the change but I think it will be better to fix this upstream, as xr.open_zarr()
is pretty heavy (i.e., all metadata is passed) just for getting the names of the data variables.
I'll open a couple of issues in the xpublish repo.
I'll open a couple of issues in the xpublish repo.
👍
Follow-up from #2.
Extract tracks query: