### 4 Terminology

An astronomical catalogue is basically a table of values, consisting of the measurements of the same property for a set of objects, together with the auxiliary information necessary to describe this table. There are several different terminologies for describing the elements of such tables. For simplicity, in CURSA a terminology which corresponds loosely to that used intuitively for the paper versions of astronomical catalogues is used:

row
the values for all the properties associated with some particular object,
column
the value of a single property for all the objects in a catalogue,
field
the value of a single property for a single object (that is, the intersection of a row and a column).

Some of the other terminologies are shown for comparison in Table 12.

 CURSA Fortran Relational Database table file relation row record tuple column field attribute field data item, field component format format schema number of columns number of fields arity, degree number of rows number of records cardinality

Table 1: Alternative terminologies for the components of tables

In CURSA each catalogue can contain only one table and the two terms can usually be used interchangeably without introducing any ambiguity. However, where it is necessary to differentiate between the two sorts of entities, table is used to denote the simple matrix of rows and columns and catalogue is used to denote the combination of a table and its associated auxiliary information. (Note, however, that this usage implies nothing about the contents of the catalogue; it may contain a published astronomical catalogue, a set of private astronomical results or, indeed, data which are entirely non-astronomical.)

A CURSA catalogue which contains celestial coordinates in a restricted format which CURSA can interpret is called a target list. The applications which convert between celestial coordinates and plot finding charts operate on target lists. Target lists are described in Section 7.

Columns may either be scalars in which case each field comprises a single datum, or vectors, one-dimensional arrays where each field comprises a one-dimensional array of values.

Columns have a number of attributes, such as their name, data type and units. A column’s attributes hold all the details which define its characteristics. The more important column attributes are described in Section 4.1, below.

Catalogues can also contain auxiliary information which applies to the entire catalogue. CURSA recognises two types of auxiliary information: parameters and textual information. A parameter is a single datum, such as the epoch or equinox of celestial coordinates stored in a catalogue. CURSA parameters are similar to FITS keywords (in fact, CURSA interprets named keywords in a FITS table as parameters). Parameters have attributes similar to columns.

Textual information is information, usually descriptive, associated with the catalogue and intended to be read by a human. For a FITS table the textual information is basically the contents of any ‘COMMENTS’ and ‘HISTORY’ keywords3.

In the jargon of relational database systems auxiliary information is often called metadata. In the context of CURSA the metadata for a catalogue comprises the details of the columns (name, data type, units, etc.), the parameters and the textual information.

#### 4.1 Column attributes

In order to use CURSA you do not need to know the details of all the attributes of a column, but there are a few which you will probably encounter. These attributes are listed in Table 2 and are described briefly below.

 Attribute Comments NAME Name of the column DTYPE Data type DIMS Dimensionality: scalar or vector SIZE Size (number of elements) of a vector UNITS Units of the column EXFMT External display format COMM Comments describing the column

Table 2: Attributes of columns

NAME  The name of the column. The rules for column names are as follows.
• The name must be unique within the totality of parameters and columns for the catalogue. This condition is necessary in order that a component (parameter or column) may be identified unambiguously when its name is used in an expression (see Appendix A).
• A name may comprise up to fifteen characters. This value is chosen for consistency with HDS and is adequate for FITS tables.
• The name can contain only: upper or lower case alphabetic characters (a-z, A-Z), numeric characters (0-9) and the underscore character (‘_’). Note that lower case alphabetic characters must be allowed in order to access existing FITS tables. However, corresponding upper and lower case characters are considered to be equivalent. Thus, for example, the names: HD_NUMBER, HD_Number and hd_number would all refer to the same column.
• The first character must be a letter.
DTYPE  The data type of values held in the column. CURSA supports the standard data types of Fortran 77 (apart from the COMPLEX data types) and also signed one and two byte INTEGERs. DIMS  The dimensionality of the column: scalar or a vector. SIZE  If the column is a vector this attribute contains the number of elements in the vector. If the column is a scalar it is set to one. UNITS  The units in which values stored in the column are expressed. The UNITS attribute is used to identify, and control the appearance of, columns of angles (see Appendix B). Apart from this exception the units are treated purely as comments and no attempts are made to automatically propagate and convert units in calculations and selections. EXFMT  The format used to represent a field extracted from a column for external display by xcatview (see section 11) or catview (see section 12). The external format specifier should be a valid Fortran 77 format specifier for the data type of the column. COMM  Explanatory comments describing the column.

2This table is adapted from Database Systems in Science and Engineering by J.R. Rumble and F.J. Smith[24], p158.

3This statement is something of an over-simplification. See Appendix C for a complete description of the way that FITS headers are interpreted as textual information.