Make it possible to choose output rebinned workspace type

[Owen Arnold]

Should be able to rebin and either generate a MDHistoWorkspace or a MDEventWorkspace. Similar to the concept of creating a DND object in Horace either with or without the backing data.

Reported by T. Perring/ R. Ewings as an issue. As this will help take slices forward for analysis once chosen in visual manner.

Should have this option available while performing a visual rebin.

[Owen Arnold]

This is partially done under my private branch VisualRebinToAnything, but completing it requires changing all the vtkDataSet factory concrete types to have additional IMDIterator construction routines. This close to the release I am nervous about making a change, so this issue will be picked up again in the next iteration.

[Owen Arnold]

refs #4643. Additional visualisation options

Changeset: fa4e3f111494a091b717575cb3177c2c7d450bab

[Owen Arnold]

refs #4643. Expose additional functionality via iterator

Changeset: f03ff96336b661e0cff6755f6b460401bca3d008

[Owen Arnold]

refs #4643. Fix include path error.

Changeset: 3030d7244d7171d80659ab0595ee391ca9d38623

[Owen Arnold]

refs #4643. Fix GCC unknown type issue.

Changeset: 4c3491769c2ae940191f79dbf6adeffe890511dc

[Owen Arnold]

refs #4643. Fix GCC unknown type issue and tests.

Changeset: 066500c4af74757ee026e111ae7da619e479e05d

[Owen Arnold]

refs #4643. Another include fix.

Changeset: 598d2a8bb9f4db3618671f92cc1524384bbb48e1

[Owen Arnold]

refs #4643. Let SliceMD choose Recursion Depth

Changeset: b1e0f7b107ed203000e80089abad590ddbcb49a1

[Owen Arnold]

refs #4643. More Tests

Changeset: ab7778555d4a16a38e10a90cf01aef7c01943bb9

[Owen Arnold]

refs #4643. Constrain iteration depth.

Also Better internal visualisation size estimation.

Changeset: 6dcbc021e73164e6d9fce1eb3c994b6573cc2ff4

[Owen Arnold]

refs #4643. Fixes dimension expansion issue.

Resulted in duplicate mappings.

Changeset: 288f14b88c0bd108fe082f68ac60763aa9bc88a6

[Owen Arnold]

refs #4643. Fix annoying switching between Histo and MDEW outputs.

Changeset: 5e62c273e577849e0643b2f9f2a778e5c55c7ac4

[Owen Arnold]

refs #4643. Benchmark vtkMDQuadFactory via new performance test.

Changeset: 9cde415578fb05c9fc26d5c2289ee91a8780141d

[Owen Arnold]

refs #4643. Going over twice as fast now.

Changeset: 5a0f39a9f450874c0fd6964f091936efbcfeb67e

[Owen Arnold]

refs #4643. 1D vtkDataSet Factory added.

Changeset: 0a28015246bfbb87303794ca47ae9bc83cf7883c

[Owen Arnold]

refs #4643. swap in new vtkDataSet factories.

Changeset: 9603dc800e8ace18bca8155b961ff09995c3639a

[Owen Arnold]

refs #4643. Fix 3D + 4D mdew vtk factory to allow chaining.

Changeset: 0dfe86742f5887ce5ee1cc03a359d0a430d50b76

[Owen Arnold]

refs #4643. Swap in updated MD 3 and 4D vtkFactory.

Also add features to cause dimensionality checks to be ignored. Changeset: c482fbb5f5d513b517b946e3dcc4450fffd21f52

[Owen Arnold]

Hopefully I'm not speaking too soon. But I think this ticket is complete now.

Tester:

• Notice that the Rebinning Filter now has a checkbox for outputing histo workspaces (which should be on by default)
• In mantid plot you should be able to toggle this on and off and hit 'Apply', which should result in no change to the image drawn. It is actually redrawn, but the image should be identical. You should however notice that the output workspace *_visual_rebin toggles from being type MDHistoWorkspace to MDEventWorkspace.
• From the logs you should see that BinMD is being run when the checkbox is checked and that SliceMD is being run when the checkbox is not checked.
• This should also work when you are collapsing dimensions down too.

[Michael Reuter]

• I see the checkbox on the filter.
• Yes, I see the behavior.
• I see the correct algorithm being run.
• This works too.

[Stuart Campbell]

This ticket has been transferred to github issue ​5490