Implement a Spectra class with an ISpectra interface

[Martyn Gigg]

In algorithms we frequently use the workspace index to access many things about a spectra and reuse the index a lot. A cleaner approach would be to have a spectra promoted to a class-level concept so that ws->getSpectrum(i) would return an object that has a reference to the workspace and the current index and could then access everything without the need for remembering the index.

It should have an interface so that the EventList can implement it as this is a broadly similar concept.

[Nick Draper]

"New" tickets moved at the code freeze of iteration 29

[Janik Zikovsky]

(In [13047]) - Refs #3203: Implemented ISpectrum. It is the base class of EventList and Histogram and contains the spectrum number and list of detector IDs for a given spectrum.

• ISpectrum return X/Y/E vectors. EventList, e.g. calculates the histogram on the fly.
• Moved the EventWorkspace MRU list of histograms to be contained within the EventList; better code separation. Made a EventWorkspaceMRU class to encapsulate this more easily.
• Workspaces must implement getSpectrum (const/not const) to return a ISpectrum object. dataX(i), etc. methods on MatrixWorkspaces no longer need to be overridden since they call getSpectrum()
• Got rid of EventWorkspace::getEventListatPixelID() and doneLoadingData()
• replaceSpectraMap() methods offers backwards-compatibility with SpectraDetectorMaps by setting the specNo and detector list when called. Must be called AFTER creating all the spectra in a workspace.
• Refs #3318: maskWorkspaceIndex() implemented as part of ISpectrum.

[Janik Zikovsky]

(In [13051]) Fixes #3321: deprecated DiffractionFocussing1. Fix tests in release version (refs #3203). Fix a few warnings Refs #3302.

[Janik Zikovsky]

(In [13052]) Refs #3203: Fix occasional segfault in ConvertToMatrixWorkspaceTest due to a little error in CRITICAL block names for OpenMP thread safety. One character fix :)

[Janik Zikovsky]

(In [13065]) Refs #3203: Test fix attempts.

[Janik Zikovsky]

(In [13067]) Refs #3203: Temporarily disable test

[Martyn Gigg]

In [13077]:

Fix the two system tests. The Aziz one just has a new reference result as the internal order of the grouped detectors has changed but this should not matter since we don't ungroup afterward. The TOSCA test failed because ExtractSingleSpectrum was pulling out the incorrect spectrum after the ISpectra changes. The blame here is due to a bad unit test, it only tested the data and not the underlying spectra/detector links. This test has been improved. Refs #3203

[Russell Taylor]

Janik - please can you find a way to eliminate the following Intel compiler warnings, which (because of the classes they come from) appear many times in the Mac build:

EventList.h:67, Intel, Priority: Normal

overloaded virtual function "Mantid::API::ISpectrum::dataX" is only partially overridden in class "Mantid::DataObjects::EventList"

EventList.h:67, Intel, Priority: Normal

overloaded virtual function "Mantid::API::ISpectrum::setX" is only partially overridden in class "Mantid::DataObjects::EventList"

EventWorkspace.h:47, Intel, Priority: Normal

overloaded virtual function "Mantid::API::MatrixWorkspace::setX" is only partially overridden in class "Mantid::DataObjects::EventWorkspace"

[Janik Zikovsky]

In [13129]:

Refs #3203: Hopefully reduce Intel compiler warnings.

[Janik Zikovsky]

In [13245]:

Refs #3203: Continuing to deprecate SpectraDetectorMap

[Janik Zikovsky]

In [13247]:

Refs #3203: Deprecating away

[Janik Zikovsky]

In [13251]:

Refs #3203: Replaced calls to getIndicesFromSpectra() where possible.

[Janik Zikovsky]

In [13252]:

Refs #3203: Updated a few more algos.

[Janik Zikovsky]

In [13253]:

Refs #3203: More spectraMap uses going away

[Janik Zikovsky]

In [13254]:

Refs #3203: Fix.

[Janik Zikovsky]

In [13255]:

Refs #3203: More updates.

[Janik Zikovsky]

In [13256]:

Refs #3203: Build fix

[Janik Zikovsky]

In [13257]:

Refs #3203.

[Janik Zikovsky]

In [13414]:

Refs #3203: Performance test comparing ISpectrum and SpectraDetectorMap

[Mathieu Doucet]

In [13435]:

Removed confusing message. Users don't need to see this. This type of messages is only useful when a solution is also given. Re #3203

[Janik Zikovsky]

In [13534]:

Refs #3400, Refs #3203. Fix for ManagedWorkspace2D and its variants. Workspace2D now holds a vector of pointers that should keep things consistent between it and its managed variants. ManagedHistogram1D is now the replacement in the managed cases.

[Janik Zikovsky]

In [13536]:

Refs #3203. Refs #3380: Removing more calls to spectraMap()

[Janik Zikovsky]

In [13555]:

Refs #3203: Removed a few more calls to spectraMap().

[Janik Zikovsky]

In [13558]:

Refs #3203: Nearly all calls to spectraMap() are gone.

[Janik Zikovsky]

spectraMap() is very nearly gone. Nick says he wants to leave it for a little while, but it is called exactly twice in all the code.

[Russell Taylor]

In [13584]:

Revert last change for now because the new way is far too slow (~100 times) in the case of a large number of detectors. Re #3203.

[Russell Taylor]

In [13598]:

MaskDetectorsInShape performance test. Re #3203.

[Janik Zikovsky]

In [13677]:

Refs #3203: Spectrum fix for CrossCorrelate

[Martyn Gigg]

In [13678]:

Refs #3203 Change from adding specNo to the detectorIDs in the EventLists. For SNS this makes no difference as they are always 1:1, for ISIS it fixes the file loading.

[Vickie Lynch]

In [13679]:

Refs #3203: Spectrum fix for CrossCorrelate

[Roman Tolchenov]

Confirm implementation of ISpectra

[Stuart Campbell]

This ticket has been transferred to github issue ​4050