The report file (named by the report=
parameter, and defaulting to astrom.lis
) is a human-readable
report of ASTROM’s actions and its results. The summary file, which normally appears on the
terminal, contains observations and warnings about ASTROM’s progress, and should be
monitored. Neither of these files, however, is intended to be easily machine-parseable, and
so if ASTROM is used within a larger system, the other output files will be useful. The
log file is a report on ASTROM’s actions and results which is easy to parse, and easy to
extract information from. The WCS files contain ASTROM’s results in the form of a FITS
header, containing FITS WCS keywords. Both these files are described in the subsections
below.
The log file is a report on ASTROM’s actions and results which is easy to parse, and thus easy for another program, acting as a harness for ASTROM, to find out about ASTROM’s progress.
The log file consists of a number of statements describing the fits ASTROM has performed, and
its success. The statements describing each fit are bracketed in a pair of statements ‘FIT
〈n〉’ and ‘ENDFIT
’, for
some fit number 〈n〉.
The statements thus bracketed are in table 1.
Result | RESULT keyword value |
Status | STATUS [
|
Residuals | RESIDUAL source-number dx dy dr |
Information | INFO code details… |
Warnings | WARNING code comment |
Errors | ERROR code comment |
The INFO
, WARNING
and ERROR
statements share a common set of codes, and are described in
Appendix D.
The RESIDUAL
message reports the size of the residuals for the given source number in the input file. The residuals
are reported in the x
and y directions,
plus δr=√δx2+δy2.
The STATUS
message is OK
if the fit succeeded, and BAD
otherwise.
The RESULT
statements give feedback about the results of the fit. Much of the information is
also in the FITS WCS headers, if they are generated. The keywords are listed in table 2.
keyword | meaning |
nstars | number of ref stars |
xrms | RMS errors in fitted x, arcsec |
yrms | …in y |
rrms | …in r |
plate | (mean) plate scale, arcsec |
prms | rrms in pixels |
nterms | number of terms in fit |
rarad | projection pole RA, radians |
decrad | projection pole Dec, radians |
rasex | projection pole RA, sexagesimal |
decsex | projection pole Dec, sexagesimal |
wcs | name of FITS WCS header file |
RESULT
statementIf requested (by the presence of the fits=
specifier on the command line), ASTROM will write out the
plate solution in a series of FITS files, containing headers conforming (largely) to the FITS WCS
standards (known as ‘Paper I’ and ‘Paper II’) [1, 2]. ASTROM will generally attempt more than one
fit. The file names will start with the string given in the fits=
parameter.
There is more than one way to encode the required WCS information, and which way is used depends
on the value of the parameter wcsstyle=
on the command line. The allowed values of this are qtan
and xtan
, and these are discussed now.
There is a standard for specifying world coordinate systems in FITS files [2], and ASTROM conforms to this. At present (May 2003) there is only an early draft standard for representing distortions, Representation of distortions in FITS world coordinate systems, Calabretta et al. (also known as ‘Paper IV’), available at Mark Calabretta’s web pages [3]. Part of ASTROM’s function is to determine and report such distortions, but since there is not yet any standardised way to do this, we have something of a problem.
This program does not attempt to produce output using the distortion model described in Paper IV;
that seems premature. Instead, it describes distortions using the model described as ‘distorted
gnomonic’ (TAN
) in the late draft versions of Paper II. If you specify wcsstyle=xtan
(not
recommended), then ASTROM emits FITS headers which fully conform to these drafts; if you specify
wcsstyle=qtan
, the FITS headers are essentially the same, but with the draft paper’s PVj_m headers
replaced by non-standard QVj_m. In this latter case, the headers are conformant with the final Paper II,
but the distortion information is available to software which knows how to use it. You are strongly
advised not to produce new FITS files using option wcsstyle=xtan
, unless you are obliged to by old
versions of software.
The ‘distorted gnomonic’ TAN
projection is not documented here (deliberately), and the drafts
describing it are no longer readily available on the web. However, should you need to, you would
probably be able to find a copy through Mark Calabretta’s WCS web pages [3].
This is an interim solution. It’s anyone’s guess how long it will take for the FITS community to agree on a final version of Paper IV. Once that is finalised, however, it’s quite possible that ASTROM’s support for the above header styles will be removed.