Update data fetcher gdac/erddap to retrieve BGC variables (oxygen)

[RaphaelBajon]

Trying to implement BGC Argo floats on argopy using the erddap server. The first goal of this request was to get oxygen variable. Yet it would be better if we could get all the variables of the BGC Argo floats (not only oxygen), loading them directly with argopy (really faster) once the new variables are implemented in the argopy source code.

[gmaze]

Thanks @RaphaelBajon for this work !
This should benefits a lot of people 👍🏻

[gmaze]

About CI tests

Every time you push new commits, CI tests are triggered here: https://github.com/euroargodev/argopy/pull/202/checks
All tests must pass ✅
And you should make sure that new features are tested.

Fetching BGC data from the erddap is already coded in tests, but skipped.
All you have to do to include tests is to fix the name of the BGC data set id from the erddap server here:

argopy/argopy/tests/__init__.py

Line 64 in 13a180b

DSEXISTS_bgc = erddap_ds_exists(ds="ArgoFloats-bio")

[gmaze]

I don't understand why CI tests with BGC dataset are successful with the standard user mode, only expert mode should possibly work 😨
There is possibly something wrong and misleading in the design of the tests, I'll check this asap

[RaphaelBajon]

'PARAMATER_DATA_MODE' is not well implemented when requesting it from the erddap server. Each biological variables must have one mode 'A', 'R' or 'D' looking at the BGC documentation (for example, Section 4.2.3 or in the Argo Manual Section 2.6.5 PARAMETER DATA MODE)
We must wait from an update of the data on the erddap server before continuing implementing the different BGC variables.

[github-actions]

This pull request was marked as staled automatically because it has not seen any activity in 90 days

[github-actions]

This pull request was marked as staled automatically because it has not seen any activity in 90 days

[SBS-EREHM]

Hello all. Status bgc support? No errors for dataset='phy'

argopy.set_options(src='gdac', dataset='bgc', mode='expert')
ArgoSet = ArgoDataFetcher().region(region_date)
df = ArgoSet.to_dataframe()

returns error in fetchers.py:556

ValueError: Errors happened with all URLs, this could be due to an internal impossibility to read returned content.

[RaphaelBajon]

Hey @gmaze,
I have made the different updates on my branch and merge the remote branch. Only 1 old pytest from test_ref_region was failing. Now, I am not able to run the pytests since I have merged the remote changes (issue accessing the data I think).
We can discuss tomorrow if you have free time for the last commits and unit tests. Zoom will be great (available all day except 14h00-15h00).

pytest -ra -v -s --durations=10 test_fetchers_facade_data.py test_fetchers_data_erddap.py test_fetchers_data_gdac.py test_xarray.py.

[gmaze]

Hi @RaphaelBajon
Awesome !
I can't today, what about Wed. 10th in the morning ?

[github-actions]

This pull request was marked as staled automatically because it has not seen any activity in 90 days

[gmaze]

Hi @RaphaelBajon !
I close this PR since it has been superseded by work done in https://argopy.readthedocs.io/en/latest/whats-new.html#v0-1-14-29-sep-2023
I'll keep the branch though, because some of your work is still relevant for more BGC support
Thanks a lot for your work, we'll find a way to acknowledge this in a future release !
gm

[gmaze]

let's keep it as a draft !