This shell script reads a three-dimensional spectral-cube NDF and presents you with a white-light image of the cube. You can then select an $X$-$Y$ position using the cursor. The script will extract and display this reference spectrum. You will then be prompted to specify various fitting parameters, such as the peak position, using the cursor. The script will then attempt to fit the emission line. The fit will be displayed and you are consulted regarding the goodness of fit. If you consider the fit to be good enough, the script will attempt to perform similar fits to all spectra within the cube, building a two-dimensional NDF image of the velocity of the line. These will use the same initial parameters as the reference spectrum, unless option -a is selected. You may view this image drawn with a key (option -d), and overlay a contour plot (with a key) of the white-light image (option -c).

If you do not force the fit to be considered ‘good’ by using the -f command-line option, the script will offer the opportunity to manually refit the spectral feature for individual pixels, such as those that were unsuccessfully fitted by the automatic procedure. In this case the velocity map will be plotted and replotted after the new fit, regardless of the -p option.

• -a
  Requests that each fit may be inspected then approved or re-fit, not just the initial reference fit. A re-fit will change the initial parameter guesses for subsequent fits, so it is recommended that you note the co-ordinates of spectra to re-fit and tackle these individually in the final manual re-fit stage. [FALSE]

• -cnumber
  Number of contours in the white-light image. Set to fewer than 1 means no contours are overlaid. [15]

• -ciindex
  The palette colour index of the contours. It should be an integer in the range 0 to 15. It is best to choose an index corresponding to white, or black or another dark colour to make the contours stand out from other elements of the plot. 0 is the background colour. KAPPA:GDSTATE will list the current palette colours. [0]

• -f
  Force the script to accept the first attempt to fit a gaussian to the line. This is a dangerous option; if the fit is poor, or unobtainable the script may terminate abruptly if it is forced to accept the fit. Additionally this will supress manual re-fitting of bad pixels at the end of the run of the script. [FALSE]

• -ifilename
  The script will use this as its input file, the specified file should be a three-dimensional NDF. By default the script will prompt for the input file.

• -lfilename
  The name of an text log file containing the fitted Gaussian coefficients for each spatial pixel. The file is written as a Starlink Small Text List (STL) described in SUN/190). The STL file comprises a schema to locate and describe the columns, and store global properties; and a formatted table of the coefficients. The schema includes the units and a brief description of each column, and the name of the input NDF used. The table lists the Gaussian centre, peak height the FWHM, and integrated flux, each with its fitting error.

• -ofilename
  The filename for the output NDF of the velocity map.

• -p
  The script will plot the final image map to the current display as well as saving it to an NDF file. It will additionally overplot the white-light image as a contour map for comparison. [FALSE]

• -rnumber
  Rest-frame spectral unit of the line being fitted.

• -ssystem The co-ordinate system for velocities. Allowed values are:
  

  "VRAD"

  – radio velocity;

  "VOPT"

  – optical velocity;

  "ZOPT"

  – redshift; and

  "VELO"

  – relativistic velocity.

  If you supply any other value, the default is used. ["VOPT"]

• -v
  The script will generate a variance array from the line fits and attach it to the velocity-map NDF. [FALSE]

• -z
  The script will automatically prompt to select a region to zoom before prompting for the region of interest. [TRUE]
  +z
  The script will not prompt for a zoom before requesting the region of interest. [FALSE]

• The velocity-map display scales between the 2 and 98 percentiles. The map uses a false-colour spectrum-like colour table so that low-velocity regions appear in blue and high-velocity regions appear in red.

• CURSA:CATCOPY may be used to convert the STL log file (see the -l option) to FITS format for analysis with the likes of TOPCAT, provided the STL has the .txt file extension. If you want just the tabulated data for your own favourite tool, the schema can be easily removed manually, or with sed excluding the lines up to and including the line beginning BEGINTABLE.

This script invokes a collection of A-tasks from the Kappa and Figaro packages.