Notes from Bright Object Engine Design Review 8/29/00

Please let me know if I've satisfactorily articulated the concern
on CR-SPLITs by Friday so I can get together with Ron next week
to talk about it.


Global Count Rates

   The first big issue that came up was a discussion on checking global
   count rates.  After careful re-reading of Ron's requirements document, I
   see that he includes them and I was mistaken about what is there.

   The checks on global count rates will be included and counted over the
   circular area that best fits the macro aperture.  Thus, may get false
   reports of health and safety violations.


Multiple Catalog Support

   The idea of allowing the bright object checker to use more than one
   catalog came up.  The Analyzer could then have smarts to choose entries
   for a particular object from one catalog over other catalogs.  The 
   could be considered for future implementations.


Reanalyzing

   The design original tried to incorporate the ability to re-analyze an
   area on the sky in a cheaper way than redoing the whole analysis if just
   the exposure time changed.

   The original concern was the time to do count rate calculations which
   are time consuming with the ETCs.  However, for HST, we will be using
   lookup tables instead which will run very quickly.

   The bottleneck is then the GSC2 query.  A compromise to the design would
   be to circumvent the GSC2 query instead of the computation of count
   rates.  The Checker will then have check methods which will take an
   array of Targets or a previous FieldAnalysis as alternative to
   a circular area on the sky.


CR-SPLIT

   There was concern about using CR-SPLIT=2 for the STIS lookup table and
   wondering if we should be using CR-SPLIT=1 instead because CR-SPLIT=2
   can be derived from two CR-SPLIT=1's, but it doesn't make sense to
   try and do it the other way around.

   

Notes About Class Details

      Check

        Need to have new methods for checking a set of targets
	or rechecking a previous FieldAnalysis with a different
	ExposureDescription:

	check(ExposureDescription, Target[])
	recheck(FieldAnalysis, ExposureDescription)


      Analyzer

	analyze method has an incorrect argument.  The first argument
	should be GOV.nasa.gsfc.sea.science.Target [], ie, an array
	instead of a scalar

	Also, the analyze method should return a FieldAnalysis object.

        Remove the reanalyze method.  The functionality will be
        replaced with the new methods on the Checker.check class.
   


      HstExposure

         Remove the AtoDGain set/get methods because AtoDGain is 
	 the same concept as Gain.

      FieldAnalysis

         Remove the recheckSeverityLevels method.  The functionality
	 will be handled through the new check/recheck methods in the
	 Checker.

	 Remove the isSignalProportionalToExposureTime method because
	 we are removing the short cuts to reanalyzing.

	 Should public add methods:

		int [] getGlobalSeverityLevelChecks()
		double getGlobalTotalSignal()
		double getGlobalTotalSingalRate()

         and corresponding protected set methods for the Analyzer to
         use. 

  Miscellaneous

     Add public set methods to ExposureDescription, HstExposure to be able
     to change the various configuration parameters.

     Add protected set methods to ObjectAnalysis and FieldAnalysis objects
     to allow Analyzer to set the results.

     We should also add a method to the FieldAnalysis to add an
     ObjectAnalysis.