by Jakob Nordin (HU Berlin)

Europe/Berlin
zoom

zoom

https://desy.zoom.us/j/81523299374 Meeting ID: 815 2329 9374
Description

Abstract:

AMPEL is a framework for streaming data analysis, written in Python, and 
with a focus on time-domain astronomy. It enables users to build 
analyses out of hierarchies of single-purpose units and coordinates the 
execution of these units on a stream of data. AMPEL can execute multiple 
independent analyses at once, de-duplicating calculations requested by 
multiple users and recording the provenance of derived data. Analyses 
are described in a static, YAML-based configuration language, and can be 
developed and tested locally before being transferred to a cluster for 
large-scale execution. While it was originally created at HU and DESY to 
filter, augment, and react to ZTF alerts, it is also being used as the 
basis for DESY’s LSST broker. Here we present the motivation for AMPEL 
and some example applications, along with some of the tools we found 
useful while building it.