Use "PyData Ecosystem" or "Scientific Python Ecosystem"?

[seisman]

In the project goals, we mention that:

Integration with the PyData ecosystem: numpy.ndarray or pandas.DataFrame for data tables, xarray.DataArray for grids, and geopandas.GeoDataFrame for geographical data.

After looking at the link https://pydata.org/, users (and even myself) may be still confused with the definition and scope of "PyData Ecosystem". The best explanation of "PyData Ecosystem" may be this post: https://stackoverflow.com/questions/18168400/the-pydata-ecosystem.

In addition to "PyData Ecosystem", a similar term is "Scientific Python Ecosystem". This term is used on the NumPy Nature paper (https://www.nature.com/articles/s41586-020-2649-2#Sec3 and the figure 2 https://www.nature.com/articles/s41586-020-2649-2/figures/2) and across the scientific Python site. The term is also well defined at https://tools.scientific-python.org/about/:

The Scientific Python ecosystem is a loose federation of community developed and owned Python projects widely used in scientific research, technical computing, and data science.

I think we should use "Scientific Python Ecosystem" instead of "PyData Ecosystem".

[yvonnefroehlich]

It’s nice that you bring up this discussion.
Actually, I was also wondering about the difference between the "PyData Ecosystem" and "Scientific Python Ecosystem" but I did not look at this in more detail. Looking at your links am I also a bit usure about the exact definition of "PyData Ecosystem" (is this a community, a conference, or a list of Python libraries / packages?).

The SPEC 0 rules we are following are related to the “Scientific Python ecosystem” (https://www.pygmt.org/dev/minversions.html) and in the AGU abstract we wrote “Scientific Python ecosystem”:

PyGMT integrates smoothly within the Scientific Python ecosystem

Despite we currently write in the README:

Integration with the PyData ecosystem: numpy.ndarray or pandas.DataFrame for data tables, xarray.DataArray for grids, and geopandas.GeoDataFrame for geographical data

[yvonnefroehlich]

So, after looking at Fig. 2 of the mentioned NumPy Nature publication (https://www.nature.com/articles/s41586-020-2649-2#Fig2) in more detail I feel using consistently "Scientific Python ecosystem" makes sense. In this figure they list different libraries and projects, which cover the data types we mention in the README:

Integration with the PyData ecosystem: numpy.ndarray or pandas.DataFrame for data tables, xarray.DataArray for grids, and geopandas.GeoDataFrame for geographical data.

[weiji14]

That sentence used to read "Integration with the SciPy stack" 😆 I suggested at #605 (comment) to change it to "PyData Ecosystem" based on this stackoverflow answer, but the intention was more around PyData "data structures" (like numpy arrays, pandas dataframes, xarray dataarrays, etc), some of which live under the https://github.com/pydata organization.

As of 2024, I think 'PyData' has expanded quite a bit since NumFOCUS is sponsoring many projects now (see https://numfocus.org/sponsored-projects/?_sft_project_category=python-interface), so the term is a bit more vague. Changing it to Scientific Python Ecosystem sounds good since we have a policy around SPEC 0 and that page has a concrete list of packages/dependencies we integrate with.

[seisman]

We will go with the "Scientific Python Ecosystem".