
Date: Wed, 15 Nov 2000 10:14:32 +0100 (MET)
From: Peter Kretschmar <Peter.Kretschmar@obs.unige.ch>
Subject: Forwarded: [Response to AI 16.8: I/O and template support for j_ical_* tool]
To: Allan Hornstrup <allan@dsri.dk>, Carol Anne Oxborrow <oxborrow@dsri.dk>, Juhani.Huovelin@astro.helsinki.fi, Niels Joergen Westergaard <njw@dsri.dk>, Sami Maisala <smaisala@astro.helsinki.fi>, Stefan Larsson <stefan@astro.su.se>
MIME-version: 1.0

Sent again, because at least CAO could not find it anymore.
The original message was from Nov 7

------- Begin Forwarded Message -------

  Hi everybody -

  fearing that otherwise this will be buried by other tasks again, I have
made a first attempt to define the framework of the j_ical_* tool, we 
discussed at the last SDAST meeting.

  Just to remind you: the basic idea was that the shadow of the collimator
in appropriate shadowgrams would be used to determine position corrections
and possibly position uncertainty information. Either a single tool or a
suitable set should be developed. To have a working base, I'm assuming 
a single tool, called j_ical_fit_collimator until someone comes with a 
better name.

* Context

  The tool is launched interactively by a JEM-X specialist, who has 
  previously selected suitable data (slews, empty fields, ...) and 
  grouped that in an Observation Group using standard ISDC tools.

  The results are studied and the selection/tool running repeated
  until the specialist trusts the results.

  Finally the results are delivered to the ISDC IC data base.

* Interfaces

  It is assumed that the actual tool runs as executable and results
  are studied with separate, standard imaging tools. Alternatively,
  the tool could be completely written as isdcroot application, combining
  data access, data processing and display functions.

  If implemented as executable, the tool would take the following data
  sets as input:
  - An OG containing suitable event data
  - A model of the collimator
  and produce the following outputs:
  - A single shadowgram combining all selected input data
    [a new data structure as this would not be in the skewed coordinate
     system of JMXi-EVTS-SHD. My proposal is JMXi-SHAD-CAL, the ending 
     being -CAL because it's produced in Calibration. The data structure
     would be a simple 256*256 image.]
  - One instance of each JMXi-CORX-MOD and -CORY-MOD
  - Position resolution information [???, see below]

  The parameter file would then contain (as a start) the following
  parameters (names open for discussion):
  inOG,     s,ql,,,,"DOL of the Observation Group"
  collMod,  s,ql,,,,"DOL of a model of the collimator"
  jemxNum,  i,ql,1,1,2,"JEM-X Instrument number (1 or 2)"
  corX,     s,ql,,,,"DOL of X position corrections (output)"
  corY,     s,ql,,,,"DOL of Y position corrections (output)"
  evtType,  i,ql,0,0,1,"Type of JEM-X event (0=Full,1=Restricted)"
  selectStr,s,ql,,,,"String used in event selection"

* Processing

  This section is mostly on SL to write, but in my view the executable
  does roughly the following:
  
  Read in parameters
  Open OG
  For each ScW in OG:
      Select events according to jemxNum, evtType and selectStr
      Read in all these events and add them to overall shadowgram
      [Note: mixing data from various ScWs should improve matters]
  Fit model shadow to shadowgram to derive position corrections
      and uncertainty estimate.
  Store output

  [PROBLEM: The information in the existing JMXi-DPOS-MOD data structure is 
   energy dependent. Do we want to do all this in several bands? If yes,
   what about the (currently independent) position correction?]


  Looking forward to your comments and ideas,
  Peter

      
             
  

  



-------  End Forwarded Message  -------

