Other domain-specific toolkits

[hammer]

It could be useful to see how scientists in other domains are solving problems using the same PyData libraries we hope to build on. I will spend some time looking at how each of these toolkits make use of NumPy, Dask, Xarray, Zarr, and other libraries to see if there are any design patterns or other experiences we might learn from.

• Astropy
  • I/O Registry may have some ideas that could inform our integration of Python IO libraries
• MetPy
• Pangeo
  • xgcm by Ryan Abernathy
  • pangeo-datastore is a nice way to make relevant data easy to find
• Open Radar Science

[hammer]

(Post by @mrocklin)

You might also want to check in with the CZI groups of Napari and Starfish, who also use the numpy/dask/zarr stack, and are slightly closer to the biological focus here

[hammer]

The Canadian Open Neuroscience Platform produced an interesting benchmark of Dask: A performance comparison of Dask and Apache Spark for data-intensive neuroimaging pipelines (2019).

[hammer]

Satpy is another.

The Satpy package is a python library for reading and manipulating meteorological remote sensing data and writing it to various image and data file formats.

[hammer]

https://github.com/pyomeca/pyomeca