Back to Search Start Over

libyt: a Tool for Parallel In Situ Analysis with yt

Authors :
Tsai, Shin-Rong
Schive, Hsi-Yu
Turk, Matthew J.
Publication Year :
2022
Publisher :
Zenodo, 2022.

Abstract

In the era of exascale computing, storage and analysis of large-scale data have become increasingly more important and difficult. In situ analysis provides a promising solution to this challenge. Here we present libyt (https://github.com/calab-ntu/libyt), an open source C++ library allowing astrophysical simulations to analyze and visualize data with yt or other Python packages in parallel during simulation runtime. We describe the methods for reading adaptive mesh refinement data structure,accessing and handling in-memory cell-centered, face-centered, and particle data with minimal memory overhead, and exchanging data among parallel processes. We provide concrete examples of incorporating libyt to a simulation code. Furthermore, we demonstrate its capability with two astrophysical simulations, fuzzy dark matter vortices and core-collapse supernovae, with the code GAMER.

Subjects

Subjects :
in situ analysis
data analysis

Details

Language :
English
Database :
OpenAIRE
Accession number :
edsair.doi.dedup.....43eba4e0a071a610e9c88ba7a40e0be1
Full Text :
https://doi.org/10.5281/zenodo.7268302