Opened 14 years ago

Closed 14 years ago

#159 closed bug (worksforme)

HPMSA and RPA code have problems with "extra" wave of information

Reported by: srkline Owned by: srkline
Priority: major Milestone: Analysis Wish List
Component: Analysis Keywords:
Cc: Blocking:
Task:

Description

The RPA code has been fixed but not checked in yet. The HPMSA code is more troublesome since when a resolution smeared version is plotted, it is coming from a "temporary" incarnation of the coefficient wave, coming from P*S. Thus the incoming coefficient wave has no information about where the original coefficient wave originated from. The x-wave is no help either, since during curve fitting, the "real" x and y are replaced with oddly-named, Igor-generated copies. Some (easy) way to keep the location is really necessary.

This may also be a problem with the XOP, too. A single static global array is used in the XOP. Is this really threadsafe?

Also, in both cases, the "temporary" waves that are actually used are the ones in root:, never the ones in the data folders.

Change History (1)

comment:1 Changed 14 years ago by srkline

  • Resolution set to worksforme
  • Status changed from new to closed

RPA currently has no XOP version, so everything is done correctly, with waves either in the root: folder, or in the data folder for the smeared calculation. No XOP will probably ever be done, as it would require a STRUCT.

I think that the HPMSA is fine as is. For the unsmeared calculation, the global wave is in the root folder. for the smeared calculations, the global wave is in the data folder. for the XOP, the globals are statically defined (derived from the input parameters) and simply used in place, no need to generate or communicate with an Igor wave.

Note: See TracTickets for help on using tickets.