My IUE software : history
- When I first dealt with IUE data for my thesis I did not use any software, nor
any computer at all ! We measured continua and even line equivalent width with
a ruler (and a calculator) from the CALCOMP plots supplied by VILSPA, or the
plots on graph paper produced at ESO (then hosted at CERN). That was 1978 !
Eventually we checked things like quality flags and line saturation on ASCII
dumps of the "5th files" produced by somebody else's programming students who
were devising some sort of reduction environment (actually I'd should say
lineprinter dumps, since they were done an EBCDIC IBM machine !).
I remember that such (never finished) system had something called "the management
of the 100 files" which is recalled in the drawing I put at the end of the
report R1 mentioned below (sorry but that is not and will never be
in computer readable form !).
- Report R1 (entry 15 in publication
list) describes the status of the earlier IUE analysis system which I set up,
with the assistance of one student, in 1980-81.
This s/w was entirely based on an IBM VM/CMS mainframe, and was fed with
the "5th files" from IUE tapes processed at VILSPA (or GSFC) or eventually
with spectra extracted with IHAP (at ESO and converted to ASCII tapes on our HP).
Some graphical analysis was possible using
the interactive Tektronix 4010 (or 4015) terminal as well as plotting on a
- Report R4 (entry 82)
describes the IUE analysis system as refurbished in 1985-87.
This was a double-headed system, with all the basic reduction from the
original IUE tapes running on the IBM mainframe, as well as the interactive
analysis previously possible and the spectral
fitting of continua, but also supporting some IHAP batches to perform
alternate data extraction from the "4th files" on the HP. It was therefore
inclusive of some bidirectional file conversion programs, and also interfaced
with the PABLO plotting program.
An addition to this system in 1989 was an IHAP batch (plus a couple of Fortran
programs) implementing the Gaussian extraction of spectra (GEX), which is
described in report R9 (entry 98).
- Report R19 (entry 120)
describes the resurrection of the IUE system (as the phoenix when its
500-th year is close as in Dante's quotation on the front page) which
survived the abandonment of the HP IHAP system in favour of the Sun based
MIDAS system, around 1991-92.
The software on the IBM remained unchanged, while the IHAP "standard" and
"GEX" extraction batches were replaced with MIDAS procedures (and Fortran
programs for GEX). It was also necessary to write a Fortran program to
convert IUE tape data to FITS for ingestion in MIDAS. The program could
run on Sun and VAX.
- See also a separate history of fitting programs
My IUE software : concepts
The de-facto concepts of this first complete package I wrote
(whose history is reported above) resulted to be :
Of these concepts the usage of a standard file format, the paradigm of the verb + file to
be acted upon (+ other arguments), the usage of interactive graphics were also
assumed voluntarily as essential for any future software.
- Use of spectra already calibrated in Flambda vs lambda (except for those
extracted from the "4th files" via the IHAP/MIDAS batches)
- Kept in formatted (EBCDIC on IBM, ASCII on HP/Sun) files with basic names
like SWP1416 FLUX, and plain tabular layout (2 columns)
- Other file types (using the same format) with different extension were used
to keep track of reduction done (i.e. SWP1416 DERED was known to
have been dereddened.
- A later format (SIGMA files) supported errors on fluxes, for spectra
obtained by broad band averages. Originally a plain 3-column tabular format,
later changed to the format used by the
PABLO plotting program
(two tabular columns, with data and errors on alternate lines), and transparent
support to both old and new format.
- Commands to perform basic operations (rebinning, smoothing, dereddening,
sum, juxtaposition, scaling etc. or interactive graphical analysis
on spectra were in the form of imperative
verbs (in Italian) with the filename as argument (filetype assumed implicitly)
e.g. RIBINA SWP1416 for rebinning, LISCIA SWP1416 for
smoothing but DERED SWP1416 for dereddening (DEARROSSA was
exceeding the 8-character name limit !)
- The program then prompted at the terminal for other arguments
(like AV for dereddening or boxcar for smoothing) : this was a limitation of
the way IBM VM/CMS Fortran handled communication which was preserved for
historical reasons although the newest programs (like the IBM fitting programs
or the Sun/VAX programs) overcame this in favour of arguments on the runstring.
- Interactive graphics was a must since the first version, which allowed to plot
a local or analytic continuum anchored using the cursor, and also to compute
interactively equivalent widths and integrals.
Lucio Chiappetti 04 Sep 02 14:25