Arxiv:2001.11604V2 [Cs.PL] 6 Sep 2020 of Trigger Conditions, Having an in Situ Infrastructure That Simplifies Results They Desire
DIVA: A Declarative and Reactive Language for in situ Visualization Qi Wu* Tyson Neuroth† Oleg Igouchkine‡ University of California, Davis, University of California, Davis, University of California, Davis, United States United States United States Konduri Aditya§ Jacqueline H. Chen¶ Kwan-Liu Ma|| Indian Institute of Science, India Sandia National Laboratories, University of California, Davis, United States United States ABSTRACT for many applications. VTK [61] also partially adopted this ap- The use of adaptive workflow management for in situ visualization proach, however, VTK is designed for programming unidirectional and analysis has been a growing trend in large-scale scientific simu- visualization pipelines, and provides limited support for highly dy- lations. However, coordinating adaptive workflows with traditional namic dataflows. Moreover, the synchronous dataflow model is procedural programming languages can be difficult because system somewhat difficult to use and does not always lead to modular pro- flow is determined by unpredictable scientific phenomena, which of- grams for large scale applications when control flows become com- ten appear in an unknown order and can evade event handling. This plicated [19]. Functional reactive programming (FRP) [19,24,48,53] makes the implementation of adaptive workflows tedious and error- further improved this model by directly treating time-varying values prone. Recently, reactive and declarative programming paradigms as first-class primitives. This allowed programmers to write reac- have been recognized as well-suited solutions to similar problems in tive programs using dataflows declaratively (as opposed to callback other domains. However, there is a dearth of research on adapting functions), hiding the mechanism that controls those flows under these approaches to in situ visualization and analysis.
[Show full text]