Proposal for a Lightweight NPL Version for Basic Earth Sciences Analysis

[negin513]

Hello,
Due to some requests and feedbacks from users, I'd suggest creating a lightweight version of the Numerical Python Library (NPL) that only include the essential packages like NumPy, Pandas, Xarray, and others necessary for Earth Sciences analysis. However, @vanderwb and I think that the main challenge here would be deciding which packages are 'essential' for Earth Sciences analysis.

To address this, @vanderwb proposed discussing this idea in the Earth Sciences Data Science (ESDS) community. By seeking input from diverse experts, we can collectively identify the critical packages.

A lightweight NPL version would bring several benefits, including faster clone times, easier maintenance for user cloning the environment, reduced disk space usage, improved compatibility, and a simplified learning experience.

I believe this issue could be resolved through a community discussion, perhaps in the form of a poll from ESDS community or a consultation period where users can suggest and vote on the packages they use the most.

I am creating this issue to track all the resources and conversations on this topic only for our reference.

[negin513]

Here is an initial list of packages that I think are mostly widely used, but this should definitely be confirmed with the user community and modified accordingly:

 - numpy
  - scipy
  - matplotlib
  - dask
  - dask-jobqueue
  - xarray
  - zarr
  - netcdf4
  - h5py
  - cf-units
  - pillow
  - pandas
  - bottleneck
  - pylint
  - wrf-python
  - ncar-jobqueue
  - nc-time-axis

[negin513]

If anyone is familiar with a similar environment (set of basic packages for Earth Sciences Analysis), please feel free to post it here so we can use it as a reference.