Data Dictionary Files

Before a table can be loaded into the database, a data dictionary (DD) must be created telling the database how to handle each column. The user may create a data dictionary themselves (either manually or using the "ddgen" program - tutorial here), or they can simply provide IRSA with enough information to create a proper DD for them.

Data Dictionary Columns:

Name Description Examples
cntr Ordinal number of the column 1, 2, 3, etc
name Name of the column as it will appear in the database, usually the same as the original name. (usually the same as originalname)
originalname The original name of the column extracted from the input table. ra, fname
description Description of the column (character string) "Degrees from the moon"
units Physical units of the column values. Leave blank if not relevant. arcsec, hours, deg, Jansky
intype Data type of the columns in the input ASCII file. Possible types include int, double, float, char, decimal, and date. int, double, char, date
format The format of data output from the database is by default identical to the format in which they are stored in the database. This field allows the format to be explicitly specified (uses C language data formatting standard). Users should always provide a format for every column in the catalog. 19s
5.3f
8.2e
10g
14d
dbtype The specific data type as stored in the database (i.e., integer, float, etc). float(63) defines a 32 bit floating point number, float(126) is double precision floating point, number(10) is an integer, number(19) is a long integer. For character data types, use varchar2(N) where N is the length of that field. float(126)
float(63)
number(10)
number(19)
varchar2(20)
nulls Boolean value that specifies whether nulls are allowed in this column y or n (default is 'y')
indx Indicates whether a column will be indexed in the database to improve query speed. The value of this parameter can be "n" or "y". Data providers should not modify this field; it will be updated by IRSA if that column is indexed. y or n (n is default)
tablFlg This parameter determines when the column appears in the query form. It may appear only in the short form ("0"), long form ("1"), or always appear ("2"). Tables that have many columns (e.g., the 2MASS XSC), are often more convenient to the user if they are separated into short (the most important or useful columns) and long (everything) forms. 0, 1, or 2 (default is 2)
groupID The groupid is, by default, the value of the column index counter cntr. If you want to "attach" certain columns to other columns, use the groupid index to attach the column to a specific index value. Note, groupid must have a value equal to that of a cntr value. Linked columns must be consecutive.

EXAMPLE: Suppose you have a column named "j_m" that corresponds to the J-band magnitude, with an index value equal to 10 and group ID equal to 10. Now suppose there is an associated error uncertainty, "j_msig" that has an index value equal to 11 and group ID equal to 11. To group these two columns together, set the group ID of j_msig to 10, thereby attaching this column to the column with index cntr equal to 10. This way, "j_msig" is always written to the output table along with "j_m". You could not link uncertainties to magnitudes in this way if the catalog had the order "j_m", "k_m", "j_msig", "k_msig", however, since the linked column must immediately follow the column it is linked to.

1, 2, etc (default is this column's cntr)
sel This boolean determines the selection status of the column. Three values are possible: "y" (yes), "n" (no), or "h" (hidden). If yes, the column is automatically selected (that is, by default the column will appear in the output table). If no, the column will not appear in the output table. If hidden, the column will not appear in the query page (i.e., it is hidden from the user). sel is set by the data provider. y, n, h
irsadef Normally equal to sel. May be set differently by IRSA. y, n, h

 

Data Dictionary Headers

Name Description
fixlen "T" refers to a fixed-length IPAC tables. "F" is non-fixed, and consequently is read sequentially.
alias To "Alias" or associate a column with a character string. This parameter is usually set to "none"
primary Primary key column. If "none", the default is "cntr"
spt_ind The column name containing the spatial index
longitude, latitude The column names containing ra and dec
x, y, z Column names of the x, y, z spatial coordinates

 

Sample Data Dictionary:

The data dictionary file 2MASS_LGA_IRSA.24808.dd corresponds to this input table from the 2MASS Large Galaxy Atlas: 2MASS_LGA_IRSA.tbl.

See the Tutorial for more information on how to generate a DD file.