Public parameter group NGFSETS
This document contains an overview of the public parameter group NGFSETS, (part of) which is included in several NEWSTAR programs.
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.
Please note that the generic prompt and help texts defined for these parameters are frequently superseded in the actual applications by alternative texts that are more appropriate to the situation at hand. Unfortunately, the present architecture of NEWSTAR precludes the inclusion of such texts in the documentation system, so the only opportunity to read that information is during program execution (in your terminal window!).
Prompt: 3rd index: channel(s) (may vary per application)
Expected input: Character *32: 1 value
Give the channel index (range) CHN of a cut-Set specification
(grp.fld.CHN.pol.iort.seq)
Remember that channel 0 is the 'continuum' channel.
Possible answers ([]=optional):
0 take the continuum channel
n1 take channel nr n1 (n1=0,1,2,3,....)
n1-n2[:n3] loop over channels n1 through n2 [step n3]
* loop over all channels for the field (wildcard)
n1-[*] loop over all channels for the field,
starting with n1 [step n3]
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: 6th index: cut(s) (may vary per application)
Expected input: Character *32: 1 value
Give the sequence-number index (range) SEQ of a cut-Set
specification (grp.fld.chn.pol.iort.SEQ).
Possible answers ([]=optional):
0 take the first cut
n1 take cut nr n1 (n1=0,1,2,3,....)
n1-n2[:n3] loop over cuts n1 through n2 [step n3]
* loop over all
(wildcard)
n1-[*] loop over all cuts for the field and channel,
starting with n1 [step n3]
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: 2nd index: field(s) (may vary per application)
Expected input: Character *32: 1 value
Give the field index (range) FLD of a cut-Set specification
(grp.FLD.chn.pol.iort.seq)
Possible answers ([]=optional):
0 take first (or only) field
n1 take field nr n1 (n1=0,1,2,3,....)
n1-n2[:n3] loop over fields n1 through n2 [step n3]
* loop over all fields in the observation (wildcard)
n1-[*][:n3] loop over all fields in the observation,
starting with n1 [step n3]
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: 1st index: grp (may vary per application)
Expected input: Character *32: 1 value
Give the group index (range) GRP of a cut-Set specification
(GRP.fld.chn.pol.iort.seq)
Possible answers ([]=optional):
0 take first (or only) group
n1 take group nr n1 (n1=0,1,2,3,....)
n1-n2[:n3] loop over groups n1 through n2 [step n3]
* loop over all available groups (wildcard)
n1-[*][:n3] loop over all available groups,
starting with n1 [step n3]
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: 5th index: iort (may vary per application)
Expected input: Character *32: 1 value
Give the interferometer or telescope index (range) IORT of a cut-Set
specification
(grp.fld.chn.pol.IORT.seq)
This index is most useful in loop specifications (parameter NGF_LOOPS).
For interferometer data, iort is an interferometer sequence number which can
not be easily interpreted. Therefore, this index is useful only in loop
specifications (parameter NGF_LOOPS). Specify a wildcard ('*') here and rely on
the SELECT_IFRS parameter for selecting interferometers.
For telescope data, iort is the telescope number (A-D being represented by 10-13). The selection possibilities are limited here and you may prefer to rely on the SELECT_TELS parameter.
Possible answers ([]=optional):
0 take first (or only) ifr|tel
n1 take ifr|tel nr n1 (n1=0,1,2,3,....)
n1-n2[:n3] loop over ifr|tels n1 through n2 [step n3]
* loop over all available ifrs|tels (wildcard)
n1-[*][:n3] loop over all available ifrs|tels,
starting with n1 [step n3]
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: Loop specifications: nr of cycles, index increment per cycle (may vary per application)
Expected input: Character *32: 1 to 16 values
With the NGF_LOOPS keyword, you may specify repetitions of the operation you are currently defining, systematically incrementing the Group, Channel, Field etc. indices for each new cycle.
This is done by specifying pairs of values:
niter1,index_incr1, niter2,index_incr2, ...
In each pair, the first value (n_iter) indicates the number of times the loop has to execute; the second value (index_incr) indicates how the Cut index is to be changed at the start of a new cycle.
Example:
The specification NGF_SETS = grp.fld.1-2.* would select the combination
of all cuts of frequency channels 1 and 2 for the field grp.fld. If one wishes
to process 32 sets of successive such pairs of frequency channels, you would
have to type in the successive NGF_SETS specifications by hand:
grp.fld.1-2, grp.fld.3-4, ....., grp.fld.63-64
Instead, you may specify NGF_LOOPS=32, 0.0.2 This will cause the program to execute the present operation 32 times in a loop, starting with the NGF_SETS specification and then incrementing its indices by 0.0.2 for every iteration; this is equivalent to the above 32 separate runs of the program.
n_iter must be > 0, and the increment can be any index string with
simple positive or negative integers. An increment of 0 may be omitted, i.e.
the increment specifications 0.0.3.0 and ..3 are all equivalent.
Loops may be nested (to a limiting depth of 8 levels). A following loop
specification is executed inside the preceding ones.
Example of nested loops:
To run your program on group 3 for 64 fields (fld index), for 10 odd
channels (chn index) per field, starting at channel 7 and combining all
polarisations (pol index) every time, specify:
NGF_SETS=3.0.7.* (initial set of cuts)
NGF_LOOPS=64,.1, 10,..2
The second loop is executed as an inner loop inside the first one, that is, for each mosaic subfield the channels are processed in a contiguous sequence.
IF YOU WANT TO BE REMINDED OF WHAT CUTS ARE IN YOUR FILE:
Type O or L in response to this prompt to get an Overview or Layout
of the file for which you need to specify the sets.
Prompt: 4th index: polarisation(s) (may vary per application)
Expected input: Character *32: 1 value
Give the polarisation index (range) POL of a cut-Set specification
(grp.fld.chn.POL.iort.seq)
POL is a number indicating the polarisation. For interferometer data,
pol=0,1,2,3 represents XX,XY,YX,YY. For telescope data pol=0,1 represents X,Y.
This index is useful for defining loops (NGF_LOOPS parameter). Otherwise you may find the SELECT_XYX parameter more convenient.
If you prefer to use NGF_POLAR here, examples of reasonable answers are:
For interferometer data
* (=XX,XY,YX,YY) 0-3:3 (=XX,YY) 0 (=XX) 3 (=YY)
For telescope data
* (=X,Y) 0 (=X) 1 (=Y)
Note: The associated NGF_LOOPS keyword allows even more looping over index values.
Prompt: Sectors to do: grp.fld.chn.pol.iort.seq (may vary per application)
Expected input: Character *32: 1 to 64 values
A NEWSTAR .NGF file contains visibilities and associated data for one
or more objects. The basic unit of data is the CUT, which is a one-dimensional
vector of some type of data extracted from a .SCN file. The data may be either
visibilities or associated data such as corrections; they may be associated
with an interferometer or with a telescope. Cuts are addressed through a CUT
INDEX which is a string of six integers separated by dots:
grp.fld.chn.pol.iort.seq
A GROUP is basically an administrative unit, allowing the user to
subdivide his data, e.g. per object.
The FIELD and CHANNEL are the field and channel numbers in the .SCN
file of the observation from which the cut was taken.
POL is a number indicating the polarisation. For interferometer data,
pol=0,1,2,3 represents XX,XY,YX,YY. For telescope data pol=0,1 represents X,Y.
Depending on the type of data, IORT is the interferometer or telescope
number. (The former is difficult to interpret; rather use the SELECT_IFRS
parameter if you want to select interferometers.)
The cut SEQuence number distinguishes cuts for which all five of the
preceding indices are identical. It is your responsibility to know what the
different seq values represent.
Index values start at zero. (Remember that for the CHN index 0 is the
continuum channel.)
You may select SETS of cuts for processing through [ranges of] values
for the ive indices, e.g.
2.3-5:2.*.1-7.*.2
The WILDCARD value '*' means 'all'. Each index may also be specified as a RANGE: <first>-<last>[<:increment>]. Indices omitted are assumed to be '*', i.e. ...1.0 means *.*.*.1.0. For wildcards at the end the dots may also be omitted, i.e. 1.0 means 1.0.*.*.*
The notation 3-5:2 stands for 'from 3 through 5 in steps of 2'. The
step must be positive. If it is omitted, it is taken to be 1 (as in '1-7'
above).
Multiple cut SETS may be specified, separated by comma's:
<Set1>,<Set2>,... The associated NGF_LOOPS keyword allows even more looping
over index values.
IF YOU WANT TO BE REMINDED OF WHAT CUTS ARE IN YOUR FILE:
Type O or L in response to this prompt to get an Overview or Layout
of your .NGF file.
IF YOU DO NOT YET FEEL COMFORTABLE WITH THESE CONCEPTS MORE HELP IS PROVIDED:
type '' or '>' to be prompted for each of the 6 indices separately,
with more specific explanation per index.
IF YOU GET BORED WITH 6-NUMBER INDICES:
Absolute Cut nrs '#<n>' can be used as an alternative.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.