Sets values for an axis array component within an NDF data structure


This routine modifies the values of an axis array component  or system within an NDF  data structure. There are a number of options (see Parameter MODE). They permit the deletion of the axis system, or an individual variance or width component; the replacement of one or more individual values; assignment of the whole array using Fortran-like mathematical expressions, or values in a text file, or to pixel co-ordinates, or by copying from another NDF.

If an AXIS structure does not exist, a new one whose centres are pixel co-ordinates is created before any modification.


setaxis ndf dim mode [comp]
file=? index=? newval=? exprs=? axisndf=?



The Data values in this NDF are used as the axis centre values if Parameter MODE is set to "NDF". The supplied NDF must be one dimensional and must be aligned in pixel co-ordinates with the NDF axis that is being modified.
The name of the NDF axis array component to be modified. The choices are: "Centre", "Data", "Error", "Width" or "Variance". "Data" and "Centre" are synonyms and selects the axis centres. "Variance" is the variance of the axis centres, i.e. measures the uncertainty of the axis-centre values. "Error" is the alternative to "Variance" and causes the square of the supplied error values to be stored. "Width" selects the axis width array. ["Data"]
The axis dimension for which the array component is to be modified. There are separate arrays for each NDF dimension. The value must lie between 1 and the number of dimensions of the NDF. This defaults to 1 for a one-dimensional NDF. DIM is not accessed when COMP="Centre" and MODE="Delete". The suggested default is the current value. []
A Fortran-like arithmetic expression giving the value to be assigned to each element of the axis array specified by Parameter COMP. The expression may just contain a constant for the axis widths or variances, but the axis-centre values must vary. In the latter case and whenever a constant value is not required, there are two tokens available–-INDEX and CENTRE–-either or both of which may appear in the expression. INDEX represents the pixel index of the corresponding array element, and CENTRE represents the existing axis centres. Either the CENTRE or the INDEX token must appear in the expression when modifying the axis centres. All of the standard Fortran-77 intrinsic functions are available for use in the expression, plus a few others (see SUN/61 for details and an up-to-date list).

Here are some examples. Suppose the axis centres are being changed, then EXPRS="INDEX-0.5" gives pixel co-ordinates, EXPRS="2.3 INDEX + 10" would give a linear axis at offset 10 and an increment of 2.3 per pixel, EXPRS="LOG(INDEX5.2)" would give a logarithmic axis, and EXPRS="CENTRE+10" would add ten to all the array centres. If COMP="Width", EXPRS=0.96 would set all the widths to 0.96, and EXPRS="SIND(INDEX-30)+2" would assign the widths to two plus the sine of the pixel index with respect to index 30 measured in degrees.

EXPRS is only accessed when MODE="Expression".

Name of the text file containing the free-format axis data. This parameter is only accessed if MODE="File". The suggested default is the current value.
The pixel index of the array element to change. A null value (!) terminates the loop during multiple replacements. This parameter is only accessed when MODE="Edit". The suggested default is the current value.
LIKE = NDF (Read)
A template NDF containing axis arrays. These arrays will be copied into the NDF given by Parameter NDF. All axes are copied. The other parameters are only accessed if a null (!) value is supplied for LIKE. If the NDF being modified extends beyond the edges of the template NDF, then the template axis arrays will be extrapolated to cover the entire NDF. This is done using linear extrapolation through the last two extreme axis values. [!]
The mode of the modification. It can be one of the following.
  • "Delete" –- Deletes the array, unless COMP="Data" or "Centre" whereupon the whole axis structure is deleted.

  • "Edit" –- Allows the modification of individual elements within the array.

  • "Expression" –- Allows a mathematical expression to define the array values. See Parameter EXPRS.

  • "File" –- The array values are read in from a free-format text file.

  • "Linear_WCS" –- The axis centres are set to the least-squares linear fit to the values of the selected axis in the current co-ordinate Frame of the NDF. This is useful for exporting to packages with limited FITS WCS compatibility and when the non-linearity is small. "Linear_WCS" is only available when COMP="Data" or "Centre".

  • "NDF" –- The axis centres are set to the corresponding Data values read from the NDF specified by Parameter AXISNDF. This is only available when COMP="Data" or "Centre".

  • "Pixel" –- The axis centres are set to pixel co-ordinates. This is only available when COMP="Data" or "Centre".

  • "WCS" –- The axis centres are set to the values of the selected axis in the current co-ordinate Frame  of the NDF. This is only available when COMP="Data" or "Centre".

The suggested default is the current value.
NDF = NDF (Read and Write)
The NDF data structure in which an axis array component is to be modified.
Value to substitute in the array element. The range of allowed values depends on the data type of the array being modified. NEWVAL="Bad" instructs that the bad value appropriate for the array data type be substituted. Placing NEWVAL on the command line permits only one element to be replaced. If there are multiple replacements, a null value (!) terminates the loop. This parameter is only accessed when MODE="Edit".
The data type of the modified axis array. TYPE can be either "_REAL" or "_DOUBLE". It is only accessed for MODE="File", "Expression", or "Pixel". If a null (!) value is supplied, the value used is the current data type of the array component if it exists, otherwise it is "_REAL". [!]


setaxis ff mode=delete
This erases the axis structure from the NDF called ff.
setaxis ff like=hh
This creates axis structures in the NDF called ff by copying them from the NDF called hh, extrapolating them as necessary to cover ff.
setaxis abell4 1 expr exprs="CENTRE + 0.1 (INDEX-1)"
This modifies the axis centres along the first axis in the NDF called abell4. The new centre values are spaced by 0.1 more per element than previously.
setaxis cube 3 expr error exprs="25.3+0.2MOD(INDEX,8)"
This modifies the axis errors along the third axis in the NDF called cube. The new errors values are given by the expression "25.3+0.2MOD(INDEX,8)", in other words the noise has a constant term (25.3), and a cyclic ramp component of frequency 8 pixels.
setaxis spectrum mode=file file=spaxis.dat
This assigns the axis centres along the first axis in the one-dimensional NDF called spectrum. The new centre values are read from the free-format text file called spaxis.dat.
setaxis ndf=plate3 dim=2 mode=pixel
This assigns pixel co-ordinates to the second axis’s centres in the NDF called plate3.
setaxis datafile 2 expression exprs="centre" type=_real
This modifies the data type of axis centres along the second dimension of the NDF called datafile to be _REAL.
setaxis cube 2 edit index=3 newval=129.916
This assigns the value 129.916 to the axis centre at index 3 along the second axis of the NDF called cube.
setaxis comp=width ndf=cube dim=1 mode=edit index=-16 newval=1E-05
This assigns the value 1.0E-05 to the axis width at index 16 along the first axis of the NDF called cube.


File Format

The format is quite flexible. The number of axis-array values that may appear on a line is variable; the values are separated by at least a space, comma, tab or carriage return. A line can have up to 255 characters. In addition a record may have trailing comments designated by a hash or exclamation mark. Here is an example file, though a more regular format would be clearer for the human reader (say 10 values per line with commenting).

# Axis Centres along second dimension
-3.4 -0.81
.1 3.3 4.52 5.6 9 10.5 12. 15.3 18.1 20.2
23 25.3 ! a comment
26.8,27.5 29. 30.76 32.1 32.4567
35.2 37.

Related Applications


Implementation Status:

Processing is in single- or double-precision floating point.