yt_xarray: Facilitating Software Reuse Between Space and Earth Sciences#
Abstract#
In this presentation, we describe recent efforts to better link two open source Python packages, yt and xarray, in order to improve the ability to use yt with NASA, cloud-hosted datasets. yt provides support for the analysis and visualization of multi-resolution volumetric data. It can read, understand, and process data from dozens of different platforms, including well-used and established astrophysical simulation data formats as well as observational and simulation data from a number of geophysical domains. Xarray is a popular metadata-preserving array library with excellent support for analysis of remotely stored data, particular through its support for cloud-native formats like Zarr. Recent work on both a new yt_xarray package and core yt has simplified the ability to analyze and visualize geospatial and geophysical datasets loaded with xarray using yt. In this presentation, we describe some of the improvements, including 3D volume rendering with embedded interpolation and use of yt analysis methods with remotely hosted data using xarray as a data backend.
A quick note on this presentation#
A collection of scripts and notebooks reproducing all the figures in this description along with the source code for generating PDF or html renderings of this JupyterBook is available at the following repository:
data-exp-lab/yt_xarray_NASA_SMD_2024
Additionally, a html rendering of this JupyterBook is available at