next up previous contents index
Next: UV Data Sets Up: MIRIAD Subroutine Library Previous: Text I/O

General Data Set Handling

  MIRIAD makes few distinctions between what some systems (e.g. GIPSY, AIPS and FITS) call ``data'' and ``header''. Instead a data ``file'' (usually called a data set in this document) consists of a collection of items. Some items are small (a few bytes) whereas others are large (e.g. the collection of pixels in an image). All items are accessed by their name (a lower case string of up to 8 alphanumeric characters. Underscore and dollar characters are to be avoided). Convention dictates the names assigned to the various items within a data set. For example, the item name ``image'' is always used to store the pixel data of an image, ``naxis'' is the number of dimensions in an image, and ``naxis1'' is the number of pixels along the first axis of an image. New sorts of data items can be invented as the need arises.

What would be conventionally call header variables, are stored as small items. The naming convention is close to the FITS standard (though the names are lower case).

The following sections describe routines that in some way package together several calls to the lower level i/o routines. Their first argument is a ``handle'' returned by one of the open routines.


next up previous contents index
Next: UV Data Sets Up: MIRIAD Subroutine Library Previous: Text I/O


Last generated by rsault@atnf.csiro.au on 14 Mar 1996