Cookbook Hackday Apr 24 2025: R tutorials with earthaccess+reticulate, or earthdatalogin?

[jules32]

We had a great Cookbook Hackday yesterday, the main topic was a Big R question when teaching: should we prioritize development in earthdatalogin R package, or since reticulate is better now, can we use that with earthaccess python package?

Eli Holmes @eeholmes showed the tutorial she is developing with PACE data. Planning something different: thinking of reticulate + earthaccess this time, something different than before when she used earthdatalogin. We are using Zarr files, which is why we’re using earthaccess since R doesn’t do well with Zarr.

@andypbarrett @ateucher also discussed: Maybe more powerful message than parallel workflows between R and Python: read and wrangle the data in python, then pass to R for the powerful statistical analysis (like Eli described in the community call on Tuesday (recording.))

The next challenge there is differences in basic data read-in across R and python - need more awareness of what the user sees/expects across these tutorials
Xarray (python) reading in, format looks as expected
Terra (R) reading in, how the structure is rendered looks garbled

To be continued, with @alexishunzinger @celiaou-podaac @battistowx , Mikala, Sheyenne, and others!