| Version 1.9 Build 1367
|
|
Up: MeasurementSet access in AIPS++
Previous: Current capabilities and design
This section considers short-term revisions to the existing MS access
classes to rationalize certain existing capabilities, and to add
missing capabilities required for application development in several
areas in the short term.
- 1.
- Consolidate Glish access layer: At present, methods to
pack output MS data into GlishRecords, and to unpack and accept input
data and selections in GlishRecord format, exist in MSFlagger,
MSRange and MSSelector. These classes do not use VisibilityIterator or VisBuffer, but use buffers of
GlishRecords to hold the data where required in C++. The proposal in
this area is to isolate the GlishRecord interface code in a separate
class called MSGlishData (or similar), and to migrate to the use
of VisibilityIterator and VisBuffer in the classes
mentioned above. This will allow consolidation of code existing
elsewhere for data averaging in time and frequency. The Glish
interface class will also be responsible for translating between a
VisBuffer and a Glish data object, which will remain in
GlishRecord representation.
- 2.
- Selection: The proposal in this area is to concentrate MS
selection and sorting only in MSSelector, including: a) add MS
selection to TaQL converter (currently in Glish); b) form union
MSCalEnums and MSSelectionKeywords; c) add sorting (col. names); d)
use new MSSelector in MSIter, imager and calibrater.
- 3.
- MSIter, VisibilityIterator, and VisBuffer: The major
changes to these classes include addition of access to selection, and
sub-table look-up. Specifically: a) add general selection to
constructor, which uses new MSSelector; b) add sub-table lookup for
current MAIN iterator interval, and propagate these changes to VisIter
and VisBuffer.
Up: MeasurementSet access in AIPS++
Previous: Current capabilities and design
  Contents
Please send questions or comments about AIPS++ to aips2-request@nrao.edu.
Copyright © 1995-2000 Associated Universities Inc.,
Washington, D.C.
Return to AIPS++ Home Page
2006-03-28