you are here: VO Desktop2008beta-doc > V Odesktop Usage > Data Analysis Cross Match
Cross Matching Examples
This page describes some techniques that can be used to cross match tables using AstroGrid
Tools are available to handle a variety of matching criteria including celestial position and time. The best method to use depends on the size of the catalogues and the desired matching criteria. The size limits are not rigid as they also depend on the number of columns and the exact method.
Matching fairly small tables
(few thousand - few 100 000 objects/rows, depending on data centre limits, file transfer and storage capacity)
- Interactive applications:
- Tasks accessible via AstroGrid: these are available via VOExplorer, e.g. in the Examples - Remote applications list.
- STILTS
- Matching two tables using STILTS in Task Runner.
- Sending a small list of objects to a table using the Coneskymatch search
- AG Cross Matcher
- AG multi-purpose Time Cross Matcher
- STILTS
- AstroGrid scripting
- Matching 3 or more tables
- Using STILTS on your desktop
- ADQL queries to multiple tables in one database - see below.
Matching small against large tables and selections from multiple tables
- Some of the options above are suitable for finding counterparts to a list of a few hundred objects or less, from a catalogue of any size up to many millions of rows. These include
- Coneskymatch in STILTS from Task Runner or in a script.
- External services such as ending a list of targets to Vizier or using the GAVO Multi-Catalogue Multi-Cone search.
- Small selection from 2 or more large catalogues e.g. all sources with a certain redshift or colour
- ADQL queries using ASTROGRID: Examples
- Some external databases allow a direct SQL query to multiple tables e.g. the WFCAM Science Archive.
- Using STILTS on your desktop: see the STILTS Help Pages. Available tasks include
- tmatch2, tmatchn for 2 or more local tables
- coneskymatch for sending a (local) short list of sources to a remote VO-compatible Cone Search service
Matching very large catalogues
(millions of objects)
- STILTS can cope with up to about a million objects but network speed (moving the catalogues to where STILTS is running) disk space, memory etc. may be the limiting factors.
- For still larger samples, please seek advice or consider whether you can refine your selection (e.g. if you only want objects within a certain flux density range or of a certain classification, you could make a selection from one catalogue first and follow one of the procedures above).
