The Program NMAP

Introduction

The NEWSTAR program NMAP makes maps (images) from the uv-data in a SCN-file, and stores them in a WMP-file. The input data may be combined in many different ways, to produce maps that range from simple to quite exotic. As a by-product, antenna patterns may be calculated (and stored in the same .WMP file), for use in NCLEAN.

The program NMAP also allows the user to manipulate maps in a .WMP file, and to convert them to other formats.

A .WMP file may contain various kinds of maps that are related in some way, i.e. different frequency channels, polarisations, pointing centres (mosaicking), antenna patterns, CLEAN residuals, and even gridded uv-data. Each image in a .WMP file consists of a 2-dimensional array of pixel values, withj a header containing descriptive information. A 'hypercube' of images in a .WMP file can have maps of different sizes. For more information about the structure of the .WMP file, see the .WMP file description.

Overview of NMAP options

The program NMAP offers the following main options:

MAKE: Make map(s) and/or antenna patterns
SHOW: Show/edit map (header) data. See section on WMP File Description.
FIDDLE: Perform all kinds of operations on maps.
W16FITS: Write FITS tape/disk with 16 bits data.
W32FITS: Write FITS tape/disk with 32 bits data.
FROM_OLD: Convert from old (R-series) format to WMP format.
TO_OLD: Convert from WMP format to old (R-series) format.
CVX: Convert a WMP-file from other machine's format to local machine's.
NVS: Convert a WMP file to newest version. This option should be run if indicated by the program).
QUIT: Quit the program NMAP.

Option MAKE: Making maps

An overview of the user interface for the MAKE option is shown in figure and its companion 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.)

Types of output images

The program NMAP produces (multiple) images of the following types:

MAP: Normal map of the uv-data or the uv-model (the uv-representation of the Selfcal model) from the SCN-file. Various linear combinations of the four measured polarisations can be specified with the keyword MAP_POLAR, to produce:
- XX, YY, XY or YX-maps: Use XX, YY, XY or YX data only
- I-map: (XX+YY)/2
- Q-map: (-XX+YY)/2
- U-map: (-XY+YX)/2
- V-map: (XY+YX)*i/2
- L-map: XX or YY or (XX+YY)/2 if both present
- *I-map: any of the above, but multiplied with
NB: Note that parallel dipoles (++) are assumed here. Observations with `crossed' dipoles (+x) require linear combinations of all four polarisations. This can be done in the map plane (see NMAP option FIDDLE_OPTION).
AP: Antenna pattern (Replace uv-data by 1's)
COS: Assume input sines to be zero
SIN: Assume input cosines to be zero
AMPL: Assume input phases to be zero
PHASE: Assume input amplitudes to be one

It is also possible to store uv-data in a WMP file, for display purposes. The uv-data is `gridded' (convolved onto a rectangular grid).

COVER: Gridded uv-coverage (data replaced by 1's)
REAL: Real part of the gridded uv-data
IMAG: Imaginary part of the gridded uv-data
AMPL: Amplitude of the gridded uv-data
PHASE: Phases of the gridded uv-data

Operations on the input data

Overview of the optional detailed-question sequences of the map-making interface. This diagram expands the QMAPS and QDATAS stubs in FIGURE .]figure.

Data selection: The uv-data that go into a map may be selected in various ways:
- One or more SCN-files.
- Sets within a SCN-file (e.g. frequ channels, pointing centres).
- HA-Scans within each Set (HA-range).
- Polarisations (XX,YY,XY,YX).
- Individual iterferometers
- An area in the uv-plane
- Clip-level
Data correction: Various correction factors are stored in the Set and Scan headers of the SCN-file. They may be applied to the uv-data at the moment that they are read from disk for processing. Application of correction is controlled by the NGEN keywords APPLY and DE_APPLY (see the section `Common Features of NEWSTAR Programs' in this Cookbook).
Data conversion: The uv-data may be converted in various ways:
- Subtraction of a source model.
- Combination of different polarisations (XX,YY,XY,YX).
- Conversion to amplitudes or phases.
Data weighting:
- NATURAL: Take each individual measured point separately, without weighting for the UV track covered by it.
- STANDARD: Weight each observed point with the track length covered on the UV plane, and average redundant baselines on a per set basis. (default)
- FULL: Weight each point according to the actual UV point density. In this case care is taken of all local UV plane density enhancements, but it necessitates an extra pass through the data.
Data tapering:
- GAUSS: (default)
- LINEAR: max(0,1-baseline/taper_value)
- NATURAL: no taper
- OVERR:
- RGAUSS:
Data convolution: (onto a rectangular grid in the uv-plane)
- GAUSS: Gaussian type with grid points
- P4ROL: Prolate spheroidal function with grid points
- P6ROL: Prolate spheroidal function with grid points
- EXPSINC: on grid points (default)
- BOX: A square box

At this point, the data are either stored in the WMP file as gridded uv-data (usually for display purposes), or Fouries transformed into an image.

Option MAKE: example

The following is an example of making a `normal' map (and its antenna pattern), using the program defaults. This is usually sufficient to get a satisfactory result. Experienced users may experiment with some of the more advanced options.

**** Put new script here ****

QMAPS: Hidden map options

The more advanced map-making options are hidden behind the NMAP keyword QMAPS. If skipped, the (context-sensitive) default values will give a satisfactory result in most cases. Their values will be printed in the NMAP log-file. For more information on each of these keywords, see the on-line Help text (type `?'), which is also printed in the `Summary of NMAP keywords' in this Cookbook.

**** Put new script here ****

QDATAS: Hidden map options

The more advanced data-selection options are hidden behind the NMAP keyword QDATA. If skipped, the (context-sensitive) default values will give a satisfactory result in most cases. Their values will be printed in the NMAP log-file. For more information on each of these keywords, see the on-line Help text (type `?'), which is also printed in the `Summary of NMAP keywords' in this Cookbook.

QDATAS (More data handling details?) = NO:

(To be added later).

Option FIDDLE: Operations on maps

Overview of NMAP's map-handling interface. } FIGURE .]{The NMAP keyword FIDDLE_OPTION offers the user a wide range of possibilities to perform oparations on maps in WMP files:

All relevant data are copied from the first (or only) input map. If 2 maps are required (ADD,AVER,POL,ANGLE) all pairs of SETS_1and SETS_2will produce an output map. SUM will average all SETS_1maps. MOSCOM will produce a single output map from all specified input maps. The others will produce an output for each SETS_1. BEAM, DEBEAM and FACTOR will overwrite the input maps. F is a specified multiplication factor.

ADD:
AVER:
SUM: Various averages of maps. In all cases the summation produces a weighted average map over all SETS_1, the weight depending on the type:
SUM: weight(i)= 1
NSUM: weight(i)= normalisation factor of map
BSUM: weight(i)= bandwidth
BNSUM: weight(i)= bandwidth normalisation factor
FSUM: weight(i)= factor given by the user. Up to 8 factors can be given, which will be used in a cyclic fashion if more needed
NSSUM: weight(i)=
QUIT: finished
POL: , unless , then 0
ANGLE: (radians), unless , then 0
EXTRACT: extract an area of maps
COPY: copy maps
BEAM: correct maps for primary beam
DEBEAM: de-correct maps for primary beam
FACTOR:
MOSCOM: combine all specified maps (referenced to same mosaic position) into one output map. The noise of the individual maps may be used as weight.
QUIT: quit the fiddle option.

Option FITS: Conversion to FITS format

There are two separate NMAP options (W16FITS and W32FITS.

newstar@nfra.nl