ANTARES Documentation

Basic Concepts

ANTARES receives Alerts and executes Filters on them in realtime. Filters crunch the data contained in each Alert and may produce data of their own, such as tags and classifications. Tags are used to direct Alerts to output streams which can then be recieved by other systems. You write Filters in Python and submit them to run live in ANTARES.

Filters have access to the following data:

  • The incoming Alert and all data contained within it.

  • The Locus object, which represents the point on the sky.

  • The history of past Alerts at the same Locus in the sky.

  • Catalog objects which match the Alert.

  • WatchLists and WatchObjects which match the Alert.

Filters can take action based on the input data. Filters can:

  • Set and update properties on the Locus. Locus properties are searchable in the database using the ANTARES website or the Client library.

  • Set properties on the Alert.

  • Set text Tags on the Locus. Tags are searchable and browsable on the ANTARES website.

Tags can be directed in realtime to Kafka streams.

The following image gives a simplified view of ANTARES:

_images/pipeline_explanation.png

Writing Filters

Get started writing Filters with the ANTARES DevKit.

Accessing Data

Query for data and connect to ANTARES’ output streams using the Client.