NAXISn

NAXISn is a set of header keywords used in the FITS (Flexible Image Transport System) format to describe the size of each axis in a multidimensional data array. The FITS header defines NAXIS, the number of data axes, and then provides NAXIS1, NAXIS2, up to NAXISN to specify the length along each axis. Each NAXISn value is a positive integer that indicates how many elements exist along that axis.

The axes are numbered from 1 to NAXIS, with axis 1 designated as the fastest-changing index in the conventional FITS data model (a Fortran-like ordering). In practical terms, NAXIS1 often corresponds to the number of columns, NAXIS2 to the number of rows, and higher-numbered axes to additional dimensions such as spectral channels or time steps. The exact interpretation of each axis can depend on the data type (image, spectrum, cube, etc.) and the convention used by the data producer.

NAXISn keywords are required for all axes when NAXIS is greater than zero; that is, if NAXIS = m, then NAXIS1 through NAXISm should be present and define the sizes of the corresponding axes. Keywords beyond NAXISm are not defined for that header. If an axis length is zero or if NAXIS is zero, the data array is scalar or empty, and the set of NAXISn keywords may be omitted accordingly.

An explicit example: a 2D image with 1024 columns and 768 rows has NAXIS = 2, NAXIS1 = 1024, and NAXIS2 = 768. A data cube with 1024x1024 spatial dimensions and 60 spectral channels has NAXIS = 3, NAXIS1 = 1024, NAXIS2 = 1024, and NAXIS3 = 60.

NAXISn is sometimes used in documentation as a generic placeholder to denote any axis-length keyword from NAXIS1 to NAXISn.