GhostCorrection algorithm for EventWorkspaces

[Janik Zikovsky]

The algorithm will take an EventWorkspace and calculate a ghost-correction histogram. It will create a new Workspace2D with the ghost correction to apply to the data.

Input Parameters:

• Original event workspace file (can be either in d-spacing units or TOF)
• Grouping/aligning .cal file (for aligning and/or DiffractionFocussing)
• Ghost Correction file (binary format file with 16 structs for each pixel)
• Rebinning parameters
• Output Workspace2D name.
• Perhaps an option to also histogram the incoming data and make another WS2D with the corrected data?

See the HistoTools SVN code for some code that does some of this.

[Janik Zikovsky]

(In [5707]) Refs #1478: GhostCorrection algorithm seems to be working well; is parallelized. There may be an issue with focussed no. of spectra not matching.

[Janik Zikovsky]

(In [5735]) Refs #1478: Added a serialized, simpler ghost correction algorithm that gives the same result as the parallelized one, but 30 times slower.

[Janik Zikovsky]

(In [5738]) Refs #1478: a std::cout command was making the build server think it was failing. It wasn't.

[Janik Zikovsky]

(In [5742]) Refs #1478: started a bit of code for a fixed version, commented out for now.

[Janik Zikovsky]

(In [5762]) Refs #1478: fix test build error and fixes to ghost correction have been started.

[Janik Zikovsky]

(In [5767]) Refs #1478: parallelized algorithm and removed superfluous option.

[Janik Zikovsky]

(In [5771]) Refs #1478: fix units of output workspace

[Janik Zikovsky]

(In [5778]) Refs #1478: spelling fix.

[Janik Zikovsky]

Still missing the second-order correction (if it is needed) but this algorithm seems to run and give reasonable results as is. Closing the ticket for this iteration; make a new one for the 2nd order correction if needed.

[Nick Draper]

fill in missing resolution

[Peter Peterson]

This works correctly.

[Stuart Campbell]

This ticket has been transferred to github issue ​2325