Private parameters for program NMAP
This document contains an overview of the parameter interface of the program NMAP. The program also uses a number of public interfaces; references to these are also listed.
Apart from a number of utility options, the main functions of NMAP are the making of maps from .SCN files (option MAKE) and manipulating these maps in various ways (option FIDDLE).
"Roadmap"-style overviews for the entire parameter interface are shown in below.
The remainder of the document describes the individual parameters in alphabetical order. This description centers on the Help texts, which have been designed to guide the user to the proper choice at each junction, even if his knowledge of the overall workings of the program is only superficial.
The first diagram, an synopsis of NMAP's parameter interface, does not yet exist. Following that diagram several sub-diagrams are planned, of which those for the FIDDLE and utility options still have to be made.
Figure shows the parameter interface for MAKE. Two detours in this interface, for parameters QMAPS and QDATAS, are normally bypassed; they are shown in figure.
Included in these diagrams are NMAP's public parameters. These are interleaved with the private ones so they can not be diagrammed as a separate unit. In fact, the only reason that they are public is that the NCLEAN 'data clean' operation uses them. This use is indicated on the right-hand side of figure and figure.
Overview of NMAP's map-making interface.
FIGURE
.]
See figure for the QMAPS and QDATS branches; by default, these branches are bypassed.
(The rightmost column in the diagram is irrelevant for map-making in NMAP.)
Overview of the optional detailed-question sequences of the map-making interface. This diagram expands the QMAPS and QDATAS stubs in
FIGURE
.]figure.
Overview of NMAP's map-handling interface. }
FIGURE
.]{
See also:
Prompt: Area: l,m, dl,dm
Expected input: Integer: 1 to 4 values
Specify the area to be selected:
l, m position in grid spacings of centre of area
0,0 is the map centre, increaing toward the upper right (i.e.
with DEcreasing RA and INcreasing DEC)
dl, dm horizontal and vertical area sizes
Prompt: Baseline averaging interval (m)
Expected input: Real: 1 value
Specify the width in metres of baseline over which visibilities will be averaged.
The minimum value is 9, representing the smallest baseline increment ever present in practice in a (set of) WSRT observation(s). The maximum is (arbitrarily) fixed at 300.
Prompt: Clipping?
Expected input: Yes/No: 1 value
'Clipping' means discarding data in a certain annulus that fall within a certain range of values (yet to be specified).parameters CLIP_AREA, CLIP_LEVELS
It is a simple (and primitive) method of eliminating data affected by strong interference. (Note that NFLAG provides a much wider scala of operations to find and suppress interference.)
Prompt: UV-radius range for clipping (m)
Expected input: Real: 1 to 2 values
Specify the (circular) UV-plane radii (in metres) between which you want to clip the data. The default is to clip everywhere.
Prompt: Amplitude range to be discarded
Expected input: Real: 1 to 2 values
Specify amplitude range (in Westerbork Units) of visibility magnitudes that you want to discard.
In the annulus defined by CLIP_AREA, values between the limits you specify will be discarded.parameter CLIP_AREA
NOTE: It would be more natural to define a range within which visibilities are considered valid. As it is, only the lower limit is actually useful, allowing you to define a rejection threshold for interference. To do so, specify your threshold for the lower and 'infinity' for the upper limit, e.g.
<threshold>,100000
Prompt: FITS comment (
Expected input: Character *70: 1 value
The given text will be included as COMMENT in FITS output.
Prompt: Convolution type
Expected input: Character: 1 value
This is the interpolation function to be used in horizontally and vertically interpolating the observed visibilities onto the rectangular grid to be used in the Fast Fourier Transform.
The choice of function determines the detailed aliasing properties of the map(s). NMAP chooses appropriate horizontal and vertical width parameters for each. You may specify one of the following functions:
Gaussian-based:
EXPSINC Sinc*exp on 6*6 grid points: An 'approximation' to the ideal
sinc (=sin(x)/x) function. This is the function selected as the
default for map-making after extensive experience with all of
the options available here.
GAUSS Gaussian type over 4*4 grid points: The function used in the
first years of WSRT operations; it was later replaced by the
prolate spheroids. The expense in computing time is the same as
for P4ROL.
Prolate spheroids: These function minimise the 'power' (= the integral of the
intensity squared) 'aliased in' from sources outside the map
P4ROL Prolate spheroidal function with 4*4 grid points.
P6ROL Prolate spheroidal function with 6*6 grid points: By using more
points in the interpolation, this function pushes the aliasing
down considerably, - at the expense of a factor two or more in
computing time for the interpolation
All the above functions may also be used in constructing a UV-plane for display (UV_COORDINATES=UV).
For plotting visibilities versus baseline or interferometer (UV_COORDINATES=BASHA or IFRHA), they are of little use, and the default one would normally select is
BOX Shift to nearest grid point.
see parameter UV_COORDINATES
Prompt: Summing multipliers
Expected input: Real: 1 to 16 values
Specify up to 8 complex weight factors by which the input maps have to be multiplied. For each factor the real part should be given first, the imaginary next. So 1,0,0,-1 means add the real part of the first map-pair to the imagingary part of the second map-pair:
(1+0*i)*(map1r + i*map1i) + (0-1*i)*(map2r + i*map2i)
If the number of weights you give is less than the number of maps to be
combined, the weights will re-used in a cyclic fashion. Example:
1,1,-1,-1 will average the first, third, ... pairs in your WMP_SETS
specification with the negated second, fourth ... pairs.
Prompt: Make line cube (Yes/No) ?
Expected input: Yes/No: 1 value
Specify if you want to output the FITS maps in one cube or in separate maps
Prompt: Action to perform on the data
Expected input: Character *8: 1 value
Specify action to perform:
SHOW show detailed map data
NOISE calculate noise
OFFSET calculate noise and offset
QUIT quit data part
Prompt: Data transformation for display
Expected input: Character: 1 value
Specify how to transform the complex input visibilities. NORMAL is the default; the others are for special experiments and diagnostics only.
NORMAL Complex value
COS Real part
SIN Imaginary part
AMPL Amplitude
PHASE Phase
Prompt: Correct map for convolution taper (Yes/No) ?
Expected input: Yes/No: 1 value
The interpolation (convolution) in the visibility domain results in a
multiplication ('tapering') of the output map(s) and antenna pattern(s) by the
Fourier transform of the convolving function; i.e., toward the edge of the map
the sources, sidelobes and grating responses appear weaker than they actually
are.
By default this effect will be corrected for by dividing the map through the taper. A side effect of this correction is that the noise, which is uniform over the whole uncorrected map, is amplified toward the map edges.
Here you are given the option to bypass this correction. e.g. because uniform noise is more important for your application than source fluxes.
Prompt: Fiddle action|
Expected input: Character *8: 1 value
Specify action to perform.
In-place modifications: Modify data of (an) input image(s):
BEAM correct map for primary beam attenuation so it will represent
the 'true' sky
DEBEAM apply primary beam attenuation to map so it will reprsent the
product of the 'true' sky and the primary beam
FACTOR multiply image with a constant factor
Unary operations: Create 1 new image from each input image:
EXTRACT extract an area from (a) image(s)
COPY copy image(s)
LOAD read or write (an) image(s) in a foreign format (e.g. Holog)
Binary combinations: make (a) new output image(s) from (a) pair(s) of input images:
ADD weighted sum of two images: F1*image1 + F2*image2
AVER weighted average of two images:
(F1*image1 + F2*image2) / [abs(F1) + abs(F2)]
POL degree of linear polarisation from Q and U maps:
sqrt (Qmap**2 + Umap**2)]
ANGLE polarisation orientation (radians) from Q and U maps:
0.5*atan (Umap / Qmap)
Combinations of more than two images:
SUM weighted summation of (a) set(s) of images in a single .WMP file
(you will be prompted to select the weighing method)
CSUM weighted summation of pairs of images. You will be prompted
for (a) set(s) of "real" images and (a) set(s) of
"imaginary" images and complex weighting factors.
RSUM idem, but the complex weighting factors will be calculated
based on a specified rotation measure.
MOSCOM 'mosaic combine': merge a set of maps (generally for different
field centres) into one output map
Miscellaneous:
QUIT Return to OPTIONS level
Prompt: Field centre: RA,DEC (decimal deg)
Expected input: DoublePrecision: 1 to 2 values
Specify the map centre wanted in the apparent-coordinate frame. Default is the fringe-stopping centre.
Prompt: l,m field shift (arcsec)
Expected input: Real: 1 to 2 values
Specify the field-centre shift in l,m coordinates.
(l,m) are 'horizontal' and 'vertical' Cartesian coordinates in a plane tangent to the celestial sphere at the reference centre. The coordinate system is B1950/J2000 or apparent as defined by your value for parameter MAP_COORD.See MAP_COORD, REF_COORD
If you enter a null value (\ or ""), you will be prompted for a FIELD_CENTRE instead. This option is intended for instrumental test programs only and has not been tested for general applications; use it at your own risk if you wish.
Prompt: Fieldsize l,m (deg)
Expected input: Real: 1 to 2 values
Specify the l and m field size of the map to be transformed. The default will produce a map with a resolution of about 3.5 grid intervals per synthesized-beam half-width.
If you give a NULL answer (two double quotes), you will be prompted for the grid steps.
Prompt: Name for output disk file
Expected input: Character *80: 1 value
Specify the file name (without an extension) to be used in creating a pseudo-tape output file name (e.g. FITS write).
Specify a full filename otherwise (e.g. LOAD/UNLOAD in FIDDLE).
Prompt: Units of source flux
Expected input: Character *4: 1 value
Specify the output units of the FITS data:
JY jansky per beam
WU Westerbork units (1 W.U. = 5 mJy)
Prompt: FFT size
Expected input: Integer: 1 to 2 values
Specify the size of the Fourier transform in the horizontal and vertical directions.
If the size in both direction is <= 17, a Direct Fourier Transform (DFT) will be made instead of the standard operation of interpolating onto a rectangular grid followed by a Fast Fourier Transform (FFT).
The standard operation suffers from 'aliasing' artefacts associated with the periodic nature of the FFT. These artefacts are suppressed, to a level that is generally acceptable, through a very careful choice of the convolution function used in the interpolstion to a rectangular grid, but they cannot be avoided completely.
By avoiding the interpolation altogether, the DFT method is free from these aliasing effects.
Prompt: Grid interval in l,m (arcsec)
Expected input: Real: 1 to 2 values
Specify the l and m grid steps in arcseconds for the map to be made. In most applications you may define GRID_SIZE as you please. NOTE however, that any number of maps that you want to combine into a single mosaic (FIDDLE MOSCOM option) must all share the same GRID_SIZE (as well as the same reference coordinates, parameters MAP_COORD and REF_COORD)See MAP_COORD, REF_COORD
Prompt: Hour-angle averaging interval (UT seconds)
Expected input: Real: 1 value
Specify the width in UT degrees of hour angle over which visibilities will be averaged (to reduce the noise per plotted point).
Note: Observations are taken at multiples of 10 UT seconds and it is
therefore convenient to specify this parameter in UT seconds as well.
The number you specify will be converted to a sidereal hour-angle
interval.
Prompt: Interferometer separation
Expected input: Real: 1 value
Specify the vertical separation in grid points between interferometers
Prompt: Input file name
Expected input: Character *80: 1 value
Specify the file name (including extension) of the file to be converted.
Prompt: Input tape labels
Expected input: Integer: 1 to 256 values
Specify the tape labels to be read. * specifies all labels on the tape.
Remember that WMP-file images are identified by indices grp.obs.fld.chn.seq). Each of the selected tape labels will be stored in the WMP-file as a separate field (FLD) in the group (GRP) being created. The CHN, POL and TYP indices will reflect the nature of the input data, SEQ will be 0.
Example:
INPUT_LABELS=3,6,8 will cause a new GRP to be created in which these
labels will be stored under the image indices
<newgrp>. 0. <chn for label 3>. <pol for label 3>. <typ for label 3>. 0
<newgrp>. 1. <chn for label 6>. <pol for label 6>. <typ for label 6>. 0
<newgrp>. 2. <chn for label 8>. <pol for label 8>. <typ for label 8>. 0
Prompt: Output map centre
Expected input: DoublePrecision: 1 to 2 values
Specify the centre of the output map in the l and m direction; in arcsec with respect to the mosaic reference position.
If you specify an *, you will prompted for RA,DEC position
Prompt: Load action/image type |
Expected input: Character *8: 1 value
Specify the type of foreign map and what to do with it:
WMP read WMP format maps from contiguous binary file
UNLOAD inverse of WMP: write WMP maps to contiguous binary file
this option can also be used to load in an SAOIMAGE file
(specify a data offset of 512)
NEWHOLOG read HOLOG file from the WSRT (this option, synonymous with
WMP, was added to help the user)
OLDHOLOG read a Holog map in old IBM-coded format (this option used to
be called HOLOG)
QUIT quit LOADing
Prompt: Map coordinate system|
Expected input: Character *12: 1 value
Specify the coordinate system for the map.
There are two choices to be made: The first is whether the map is to be made in apparent coordinates for the epoch of the observation or in fixed-epoch coordinates. The latter are fixed for each instrument: B1950 for the WSRT, J2000 for the ATNF. The second choice is whether the 'reference position' for the map must coincide with the fringe-stopping centre or is to be specified by you.
The reference position is the position at which the map plane is tangent to the celestial sphere; it defines the geometry of the map's (l,m) grid in terms of RA and DEC.
The reference position is important for mosaic mapping: The FIDDLE/MOSCOM operation that combines mosaic subfields into a single large map will only work if all input maps have the same reference.
For a single mosaic, the program by default uses the mosaic centre as the reference for all subfield maps. However, if you intend to combine multiple mosaics into a 'super-mosaic', only you can define the common reference centre that will be needed.
You have the following options for your reply:
Reference position defined by the observation (i.e. coinciding with the
fringe stopping centre):
B1950_J2000 in epoch coordinates for the epoch defined by the
instrument with which the observation was made:
B1950 for the WSRT
J2000 for the ATNF
APPARENT in apparent coordinates at the time of observation
Reference position to be defined by the user through an additional parameter
REF_COORD:
REFER in B1950_J2000 coordinates
AREFER in APPARENT coordinates. Note that this will not work for
Prompt: Input-map multipliers
Expected input: Real: 1 to 2 values
Specify the factors by which the input maps have to be multiplied.
You may specify up to 8 factors which will be used as multipliers in the weighted summation the input maps you selected. If there are more maps to be summed than factors specified, the factors will be cyclically re-used.
Prompt: Polarisation threshold (W.U)
Expected input: Real: 1 value
Specify the minimum level in Wetsrebork Units that is still to be considered valid linear polarisation. Polarisation levels below this threshold will be set to zero in the output map(s).
Prompt: Select 1 to 4 output-map polarisation(s)|
Expected input: Character *4: 1 to 4 values
Specify up to four polarisations for the maps to make:
XX XX only
XY XY only
YX YX only
YY YY only
I Stokes I
L 'line' Stokes I: Incomplete input data (e.g. no valid XX or YY)
will be filled in aassuming that the field is unpolarised
(Q=U=V=0)
Q Stokes Q
U Stokes U
V Stokes V
Each of the above may be suffixed with 'I' to indicate that visibilities must be pre-multiplied with sqrt(-1)
Prompt: Old R-series data format
Expected input: Integer: 1 value
Specify the old R-series data type:
0 local
1 VAX, D_FORMAT
2 VAX, G_FORMAT
3 ALLIANT
4 CONVEX
5 IEEE
6 DEC station
7 SUN station
8 HP station
Prompt: Action|
Expected input: Character *8: 1 value
Specify action to perform:
Primary operations:
MAKE make map(s) from visibility data in .SCN file
FIDDLE combine or change maps in .WMP file
FITS conversions:
W16FITS write FITS tape/disk with 16 bits data
W32FITS write FITS tape/disk with 32 bits data
WRLFITS write FITS tape/disk with 32 bits float data
RFITS read FITS tape/disk data
Miscellaneous:
SHOW show/edit map data
QUIT finish
Format conversions:
CVX convert a map file from other machine's format to local
machine's
NVS convert a map file to newest version. Needs to be run only if
indicated by program
FROM_OLD convert from R-series format
TO_OLD convert to R-series format
Expected input: Character *8: 1 to 8 values
Specify one or more output types:
Standard image types for map-making:
MAP Output (a) map(s)
AP Output (an) antenna pattern(s)
Visibility-domain outputs, for diagnostics only:
COVER Output the 'antenna-pattern' convolved visibilities
REAL Output the real part of the convolved visibilities
IMAG Output the imaginary part of the convolved visibilities
AMPL Output the amplitude of the convolved visibilities
PHASE Output the phase of the convolved visibilities
Prompt: Output label
Expected input: Integer: 1 value
Specify the first output tape label. If this label already exists, it and all the subsequent labels will be overwritten. Specify * or 0 to write the new label behind all existing ones.
Prompt: Output map centre
Expected input: Integer: 1 to 2 values
Specify the centre of the output map in the l and m direction in pixels with respect to the mosaic reference position.
If you specify an *, you will be prompted for l,m and RA,DEC position
Prompt: Output map size
Expected input: Integer: 1 to 2 values
Specify the size in grid points of the output map(s) in the horizontal and vertical directions.
Prompt: Special data selection (Yes/No) ?
Expected input: Yes/No: 1 value
Maps are normally made directly from the .SCN-file visibilities. Answering YES here gives you access to some specials including
- making a map from model visibilities in the .SCN file;
- selecting visibilities from an annulus in the UV plane;
- clipping extreme amplitudes in an annulus in the UV plane;
- shifting the pointing centre to which the visibilities refer (and
consequently the centre of the map to be made from them).
Prompt: Output map centre: RA,DEC (decimal deg)
Expected input: DoublePrecision: 1 to 2 values
Specify the centre of the output map: RA and DEC in decimal degrees.
Prompt: Map reference coordinates: RA,DEC (decimal deg)
Expected input: DoublePrecision: 1 to 2 values
Specify (in decimal degrees) the RA and DEC of the reference coordinates to use in producing the map. The coordinate system is B1950/J2000 or apparent as defined by your value for parameter MAP_COORD.
Prompt: Rotation measure
Expected input: Real: 1 to 64 values
Specify one (or more) rotation measure(s) (RM). Pairs of input maps will be
phase-rotated to the the frequency of the first input map and averaged:
THETAn = 2 * RM * ( (c/FRQn)**2 - (c/FRQ1)**2 )
Qout = SUM( Wn * ( cos(THETAn)*Qn - sin(THETAn)*Un ) ) / SUM( Wn )
Uout = SUM( Wn * ( sin(THETAn)*Qn + cos(THETAn)*Un ) ) / SUM( Wn )
A pair of Qout/Uout will be produced for each RM given. E.g. to get average Q/U
maps for rotation measures from 0 to 5 with intervals of 0.5, specify
ROTATION_MEASURE=0 TO 5 BY 0.5.
Prompt: Model subtraction (Yes/No) ?
Expected input: Yes/No: 1 value
Reply YES if you want to subtract a source model. You will then be prompted to provide details on the model you want to subtract.see the NMODEL HANDLE interface
Prompt: Summing multipliers
Expected input: Real: 1 to 8 values
Specify up to 8 weight factors by which the input maps have to be multiplied. If the number of weights you give is less than the number of maps to be combined, the weights will re-used in a cyclic fashion.
Example:
1,-1 will average the first, third, ... maps in your WMP_SETS
specification with the negated second, fourth ... maps.
Note: When used together with the RSUM option, the weights will be used both for the Real and Imaginary map (they are real weights).
Prompt: Weighing method
Expected input: Character *8: 1 value
Specify the type of weight to use in the averaging. In all cases the summation produces a weighted average map over all SETS_1, the weights depending on the method you select:
SUM weight(i)= 1
NSUM weight(i)= normalisation factor of map(i)
NSSUM weight(i)= 1 / (<noise in map(i)>**2)
BSUM weight(i)= bandwidth of map(i)
BNSUM weight(i)= bandwidth * normalisation factor of map(i)
FSUM weight(i)= factors to be specified by you.
QUIT quit AVERaging
Prompt: Taper type
Expected input: Character *8: 1 value
The taper is a function of baseline length used to de-emphasize the long baselines and consequently reduce the near-in sidelobes of the synthesized beam.
You may specify the following functions, some of them to be supplemented later with a baseline-scale parameter TAPER_VALUE:see public parameter TAPER_VALUE
GAUSS exp -(<baseline>/TAPER_VALUE)**2
standard WSRT beam: good compromise between near-in sidelobes,
beam width and noise
LINEAR max (0, 1-baseline/TAPER_VALUE)
OVERR 1 / baseline (no scale)
RGAUSS exp -(<baseline>/TAPER_VALUE)**2 / <baseline>
broader beam, very low near-in sidelobes, poorer noise
NATURAL no taper (no scale)
optimum signal/noise ratio, narrower beam with very strong
near-in sidelobes
NOTES: - Unless you have specified UNIFORM=NATURAL for the UV coverage mode, the
1/<baseline> density variation of measured visibility points is already
being accounted for, so OVERR and RGAUSS should not be chosen. LINEAR
does not combine very well with NATURAL either.
cf. parameter UNIFORM
Prompt: Comment to be included in map header(s) (
Expected input: Character *24: 1 value
Give, optionally, a descriptive comment for the maps.
Prompt: Visibilities to use
Expected input: Character *8: 1 value
Specify the type of visibilities to use:
STANDARD observed visibilities
MODEL model visibilities (to be specified later with type=0
sources)
Prompt: Weigh with noise (Yes/No) ?
Expected input: Yes/No: 1 value
Specify if you want the noise of the individual maps to be used as a weight in the MOSCOM combination
Prompt: UV coordinate system
Expected input: Character *8: 1 value
Specify the type of UV coordinates wanted for UV-plane type outputOUTPUT: COVER, REAL, IMAG, AMPL, PHASE options
UV standard UV coordinates: interferometer tracks are ellipses
BASHA hour-angle (horizontal) and interferometer baseline (vertical)
coordinates: interferometer tracks are horizontal lines;
redundant baselines overlap
IFRHA as BASHA, but vertical axis is the interferometer ordinal number
in the sequence 01,02,...,0D,12,13,...,CD)
Prompt: Relative weight limit
Expected input: Real: 1 value
Specifies the relative weight as compared to the expected maximum weight of data points combined on one line, below which no output will be generated
Prompt: First WMP node name
Expected input: Character *80: 1 value
Specify the node name for the first Fiddle input set of images.
Prompt: Second WMP node name
Expected input: Character *80: 1 value
Specify the node name for the second Fiddle input set of images.
Specify * if this is the same as the first node (NODE_1).
Prompt: First image set(s) to be used: grp.fld.chn.pol.typ.seq
Expected input: Character *32: 1 to 64 values
Specify the first image set(s) to be used.
Prompt: Second image set(s) to be used: grp.fld.chn.pol.typ.seq
Expected input: Character *32: 1 to 64 values
Specify the second image Set(s) to be used
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.