Documentation on contsub

Task: contsub
Purpose: Subtract continuum from a datacube
Categories: map manipulation

 CONTSUB is used to subtract the continuum from a datacube, which
 can be arbitrarily oriented. Several subtraction algorithms are
 possible, selected using the mode keyword. The maximum length of
 the frequency axis is 8192 pixels.

Key: in
	Standard in keyword. Use `help in' for more information.

Key: region
	Standard region keyword. Use `help region' for more information.
 Anything outside the region is untouched.
 (for the moment only rectangular boxes are recognized by contsub).

Key: out
	Standard out keyword. Use `help out' for more information.
 For modes poly, mean and avgs either this keyword or cont= must be
 specified, for mode subtr it must always be given. This output
 dataset will contain only those channels from the input dataset
 selected by the 'region' keyword.

Key: cont
 An optional output dataset which will be one plane deep and contain
 the continuum map. For mode=poly, this is the interpolated map at the
 channel halfway between the first and last line channel selected. For
 mode=mean, it is the averaged image that was subtracted from the
 channel maps. For mode=avgs it is the average of the two average
 images that are calculated (see description of mode keyword). For
 mode=subtr this is an input set, which will be subtracted from each
 plane of the input dataset. For modes poly, mean and avgs either this
 keyword or out= must be specified, while for mode subtr it must always
 be given.

Key: contchan
 Select channels that supposedly contain continuum signal only. A
 selection is a list of ranges. A range is specified as (z1,z2),
 meaning channels z1 through z2, or as (z3) for a single channel.
 Ranges are separated by commas. E.g.: (z1,z2),(z3),(z4,z5).
 For options 'poly' (or 'mean'), the fit (or average), is done
 using all selected continuum channels. For option 'avgs', the
 selection of continuum channels must consist of two ranges.
 The 'contchan' keyword always refers to frequency channels,
 independent of whether the frequency axis is the 'z'-axis, the
 'x'-axis or any other axis.
 (Not used for mode='subtr').

Key: mode
 Selects the continuum-determination algorithm. In each case the
 operation is done separately for each image pixel. May be
 abbreviated. The default is 'poly,1'.

 'poly,#' - a polynomial fit of order # is made through the selected
 continuum channels. The interpolated continuum intensity is
 subtracted from each separate channels. The maximum order is 8.

 'mean' - the average intensity of the selected continuum channels is
 calculated and the average is subtracted from each channel.
 (equivalent to 'poly,0').

 'avgs' - an average of the first range and of the second range of
 channels is calculated and a linear interpolation between the two
 averages is subtracted from each channel.

 'subtr' - at each position, subtract the value in the dataset given
 by cont= from the profile at the same position in the input dataset.

Key: options
 Select which output to write with the cont keyword. By default the
 continuum is written. If options=coeff,# one of the coefficients of
 the polynomial fit is written instead. '#' gives the order for which
 the coefficient is written (e.g. if cont=a+bx: 0='a', 1='b', etc).

Key: velaxis
 For datacubes where the 'VELO' axis-identification does not occur in
 the header, this can be used to indicate which direction is the
 velocity (channel) axis. The value given must be one of 'x', 'y',
 'z', 'a', 'b', .... If velaxis is not given and the 'VELO' axis is
 not present, contsub assumes 'z'.