Nearest neighbour

You are not logged in

Find the nearest neighbour to each object in
or in your catalogue

or Manual input

Neighbours found in
Within degrees (unless ' or " used)
Include all neighbours within maximum distance
Restrict to single nearest neighbour

You need to be logged in to request long running jobs and the machine-learning based match

Help

Maximum number of entries

Output format

Match area limits

Switch to equatorial

Switch to galactic (very slow, not recommended)

RA between and
Dec between and

l between and
b between and

Select columns to display

Description

For each object in the source catalogue, this will find the nearest neighbour(s) in the target catalogue, to the maximum distance specified or 2.0 degrees, whichever's smaller.

Note that when using a source catalogue from the server (ie not supplied by the user), the system will limit the search to those found on the overlap of the positional uncertainties This will probably change in the future to allow searching in a wider scope

Description and instructions about the machine-learning based matcher are below

If you choose to upload your own catalogue, the catalogue will be kept on server until you remove it. Formats understood by the system are described below

Machine-learning based matcher

The machine-learning based algorithm tries to match the same object in the catalogues given. This is done by trying to model the magnitude of one of the catalogues with a mixture of gaussians. This works only if the objects to be matched have a different magnitude profile from the background objects, and in those cases it should give better results than a plain positional match, especially in cases where the positional uncertainties are large.

If you are planning to use the machine-learning based matcher on your catalogue, make sure you have marked in the catalogues subsection of "Personal data" (here) the meaning of the columns in the catalogue. In all cases, you should mark the ID, RA, and dec columns. If the system is modelling the magnitude of your catalogue, you also need to mark the magnitude or flux column. If the system is not modelling the magnitude of your catalogue, you should mark the positional uncertainty column if one exists or the global positional error otherwise.

Upload formats

Currently the following formats are supported when uploading or typing/pasting catalogues into the text box (for all formats, the system understands coordinates in degrees only)

VOTable - preferred format. Attributing the principal RA column with a UCD of POS_EQ_RA_MAIN or pos.eq.ra;meta.main and the Dec column with POS_EQ_DEC_MAIN or pos.eq.dec;meta.main will let the system know which columns to use for RA and Dec respectively. If there is a preferred ID column, it can be attributed with a UCD of ID_MAIN or meta.id;meta.main but otherwise a counter ID column will be supplied. Alternatively, columns can be attributed with UType of pos.ra, pos.dec or id respectively.
CSV (Comma Separated Values) format - Even though it is called 'Comma' separated values, it will understand other forms of character-separated tabulation (eg columns separated by spaces). You can either:
- Arrange column order as follows: ID column, RA column, Dec column, all other columns
- Set the appropriate column names to "ID", "RA", "DEC"