refs #9257. More performance work.
Also start fixing a bug whereby the cluster offsets can become larger than the range of an integer. This has been partially fixed by better calculating the number of possible clusters that could exist. The full fix should be to also change the type of m_id in DisjointElement from int to size_t! Another issue that is being worked on here is a faster clone of the MDWorkspace. We only care about it's shape, so we just need something the right size to start putting data into!
Showing
- Code/Mantid/Framework/API/inc/MantidAPI/IMDHistoWorkspace.h 2 additions, 0 deletionsCode/Mantid/Framework/API/inc/MantidAPI/IMDHistoWorkspace.h
- Code/Mantid/Framework/Crystal/inc/MantidCrystal/ConnectedComponentLabeling.h 6 additions, 0 deletions...rk/Crystal/inc/MantidCrystal/ConnectedComponentLabeling.h
- Code/Mantid/Framework/Crystal/inc/MantidCrystal/IntegratePeaksUsingClusters.h 3 additions, 1 deletion...k/Crystal/inc/MantidCrystal/IntegratePeaksUsingClusters.h
- Code/Mantid/Framework/Crystal/src/ConnectedComponentLabeling.cpp 56 additions, 13 deletions...ntid/Framework/Crystal/src/ConnectedComponentLabeling.cpp
- Code/Mantid/Framework/Crystal/src/IntegratePeaksUsingClusters.cpp 39 additions, 4 deletions...tid/Framework/Crystal/src/IntegratePeaksUsingClusters.cpp
- Code/Mantid/Framework/Crystal/test/ConnectedComponentLabelingTest.h 7 additions, 4 deletions...d/Framework/Crystal/test/ConnectedComponentLabelingTest.h
- Code/Mantid/Framework/MDEvents/inc/MantidMDEvents/MDHistoWorkspace.h 4 additions, 0 deletions.../Framework/MDEvents/inc/MantidMDEvents/MDHistoWorkspace.h
- Code/Mantid/Framework/MDEvents/src/MDHistoWorkspace.cpp 5 additions, 0 deletionsCode/Mantid/Framework/MDEvents/src/MDHistoWorkspace.cpp
Loading
Please register or sign in to comment