A source table is used by BBS to know the available sources and their parameters. It is a special form of a Parameter Table and is described in detail in parmdb.pdf .

The sources in a source table can optionally be grouped into so-called patches. BBS makes use of patches by, for instance, combining them when doing parameter fits. It is possible in makesourcedb to define patches and tell to which patch a source belongs. If no patch is defined for a source, the source is its own patch. Thus for each such source a patch is created, where it is possible to add a prefix and/or suffix to the patch name making it possible to use patch patterns in BBS.

The program makesourcedb can be used to fill or append a source table with (selective) sources read from an ASCII file. Note that the script gsm.py can be used to extract sources from the GSM and put them into a file that can directly be used as input for makesourcedb.

It can be run like:

makesourcedb key=value key=value ...

The possible keywords and values are described in the following table.

makesourcedb reads lines from the input file, interprets them using the format string (described below), and writes the ones matching the optionally given circle or box into the Source Table. Possible comment lines (indicated by #), empty lines, and format line in the input file are ignored when reading the values from the file. Whitespace (one or more blanks and/or tabs) is ignored, unless used as separator. Leading and trailing whitespace is always ignored.
Adding a source to the source table means two things:

• it is added to the source table itself
• for all source parameters appropriate entries are added to the associated parm table. The names of those parameters look like sourcename:fieldname where field name is the name of a field given in the format string.

The format string defines which fields are present in the ASCII file and which separators are used. For example:

Name, Ra, Dec

tells that the lines in the input file contain source name, ra, and dec. It also tells that the values on each line are separated by commas. A field name can consist of alphanumeric characters, underscores and colons. A colon cannot be the first character. All other characters (including whitespace) can be used as separator. A colon can also be used as field separator as long as it is surrounded by whitespace in the format string (to distinguish it from a colon in a field name).

A few special options are available in the format string.

• Default values are used for missing or empty field values in input lines. Standard default values (as listed in the table below) exist, but it is possible to specify an explicit default value in the format string. It can be done by giving a value (enclosed in quotes) for a field like:

 Name, Ra, Dec, I='1.5'

• It is also possible to define a field with a fixed value that does not appear in the input lines like:

 Name, Ra, Dec, Category=fixed'2'

In this way the Category parameter will be defined for all sources, while the input lines do not contain the Category field.

• If the input lines contain fields that have to be ignored, they can be skipped by giving no field name in the format line or by using the field name dummy. For example:

 Name, Ra,,Dec

means that the third field is skipped. Note that the name dummy has to be used if whitespace is used as separator like:

 Name Ra dummy Dec

The format strings can be given in a few ways:

• Directly in the program format keyword.
• As a format line in the input file (before the first value line) or optionally in a separate text file. Such a format line has to start with format= (whitespace is allowed). For backward compatibility the format line can also be given as a special comment line like #(…)=format, where … is the format string. Whitespace can be used.

makesourcedb distinguishes two types of fields: standard fields and other fields. The standard fields have a predefined (case-insensitive) name. However, the name is case-sensitive in the sense that the name given in the table below is used as the parameter name in the parm table. The following standard fields are known:

All non-standard fields have type 'double' and a default of 0. The name as given in the format string is used as the parameter name in the parm table.

Note that the polarized flux can be given directly (as Q and U) or by means of the rotation measure, polarization angle, and polarized fraction. An error is given if both are given for the same source.

If LogarithmicSI is true (the default), the source intensity 'I' is computed as

I = I0 * (v/v0) ^ (c0 + c1 * log10(v/v0) + c2 * log10(v/v0)^2 + …), where 'v' is the frequency and 'v0' is the reference frequency.

If LogarithmicSI is false, the spectral index is interpreted in a way compatible with wsclean, namely as

I = I0 + c0 * (v/v0 - 1) + c1 * (v /v0 - 1)^2 + c2 * (v/v0 -1)^3 + …

This definition is also described in the WSClean documentation.

The flux of a source can be defined using a rotation measure by means of the field name RotationMeasure. The required polarization info can be given in two ways:

• Directly by means of the fields PolarizationAngle and PolarizedFraction. They define the angle and fraction for wavelength 0. If the field ReferenceWavelength is also given, it must be 0.
• Indirectly by means of the Q and U flux. ReferenceWavelength should be given as well to define the wavelength they are valid for. If not given, it will be derived from the spectral index ReferenceFrequency. The polarization angle and fraction will be calculated taking the spectral index into account like:
  polang = atan2(U,Q) / 2 - refwvl^2 * rm
polfrac = sqrt(Q^2 + U^2) / I(refwvl)
  ReferenceWavelength must be > 0 if a spectral index is given.

A patch is defined when an input line contains a patch name and no source name. For a patch the fields Category, Ra, Dec, and I are also used which define its category, center, and total flux. In a future version of makesourcedb the center and total flux can be calculated from the sources that are part of the patch.

A source is attached to a patch if its input line contains a patch name. A line defining that patch should have been processed before this line.
If no patch name is given for a source (as is usually the case), the source forms a patch in itself and the source name is used as the patch name (with an optional prefix/suffix). In this way BBS can always deal with patches.

Until 2022, the orientation of Gaussian sources has been interpreted somewhat wrong. It was interpreted (both by DP3 and pybdsf) as the angle with respect to the m axis of a certain image. The correct (implicit) definition is the orientation w.r.t. the declination axis. For fields far away from the NCP, and/or sources not too far from the field center, these two definitions are almost identical.

In DP3 v5.3 a fix was implemented to take the projection effects into account properly. This fix is only called when OrientationIsAbsolute is true. A script will be added to LSMTool to convert from 'relative' orientation to absolute ones (this script needs the phase center of the original image as an input).

Angles can be given in a variety of ways using the MVAngle format available in casacore.

• In sexagesimal HMS format like 12:34:56.78 or 12h34m56.78. Note that a colon defines hours. To avoid mistakes, it is not allowed to use colons in declination values (but see below).
• In sexagesimal DMS format like 12.34.56.78 or 12d34m56.78.
• As a double value optionally followed by a unit like 12.34deg or 12.34rad. If no unit is given, radians are used.

In all cases negative values are possible. It is checked if RA is between -360 and +360 degrees and DEC between -90 and +90 degrees.

Some catalogues use a different way to specify Ra or Dec. For this reason the special field names rah, ram, and ras can be used in the format string. They define the hour, minute, and second part of an RA. If the Ra is in degrees, rad should be used instead of rah. Similarly dech, decd, decm, and decs can be used for declination. A negative hour or degree part means that the entire value is negative. For example:

  rah ram ras

can be used in the format string to interpret RA values like:

  12 34 56.78
 -12 34 56.78

Note that these special field names can also be used to interpret declination values given in degrees using colons. For instance:

  decd : decm : decs

to interpret

     12:34:56.78

This program makes it possible to merge two SourceDB data sets or to copy (part of) a SourceDB. Patches for which the positions in both SourceDBs match (within a given uncertainty radius) are merged. It means that the patch in the second SourceDB will be used!! The mode parameter tells if non-matching patches in the second SourceDB should also be copied.

It is possible to select patches using a name pattern.

It can be run like:

mergesourcedb key=value key=value ...

The possible keywords and values are described in the following table.

This program can be used to show the contents of a SourceDB data set.

It can be run like: It can be run like:

showsourcedb key=value key=value ...

The possible keywords and values are described in the following table.