ConvertToMD: Add explicit option to convert to QLab

[Nick Draper]

The result of the ticket #6109 has shown that ConvertToMD is ready to replace ConvertToMDEvents with one small change.

At the moment the selection of QSample over QLab is implicit depending on the presenece of a UB matrix. Make this an explicit choice so users can still convert to QLab if a Goniometer is set.

One possibility would be to change the QDimensions box to contain QLab3D, and QSample3D rather than just Q3D.

If a users picks QSample3D without a Goniometer the algorithm should fail before too much processing, explaining why and suggesting QLab or ways to set up the required data.

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 strange merge with develop problems

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: f8cc80541f123b89eabbcc69e51dcb61b8849ad7

[Alex Buts]

To Tester.

Some commits erroneously went to refs #6865, so Track is not showing full picture. Despite plenty of merge conflicts with develop one can see here, merging with master should go without any problem.

I decided not to modify Q3D field, as Q3D/ModQ/CopyToMD field comes from standard Mantid factory, used for e.g. algorithm registering and reimplementing/modifying this factory would need at least big discussion.

Here I just exposed new property (Q3DFrames), which selects Q-frame (lab/sample/HKL) and added AutoSelect mode to it. Autoselect enables the previous behaviour, when lab/sample/hkl forces correspondent Q-frame. I have actually copied this property together with description from ConverToDifractionWorkspace algorithm, just renaming it for clarity and add/modified couple of unit tests to check this behaviour.

Tester should play with new property (probably from GUI) and check if the results he is getting are as expected. If he is not familiar with the results, he may enable debug mode and look at the logs for transformation matrix. The matrix should change as function of Q3DFrames selected.

The if one does not have UBMatrix (sample) attached to the workspace, and user requested HKL or sample frame, the algorithm should fail complaining about invalid arguments.

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 strange merge with develop problems

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: f8cc80541f123b89eabbcc69e51dcb61b8849ad7

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 strange merge with develop problems

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: f8cc80541f123b89eabbcc69e51dcb61b8849ad7

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 peculiar conflict on branch change (does not affect master)

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: 67cf856efca196206a00eb5bbda2787d4ac4569e

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 strange merge with develop problems

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid into develop

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h

Changeset: f8cc80541f123b89eabbcc69e51dcb61b8849ad7

[Michael Reuter]

Not going to have time right now.

[Andrei Savici]

No UB, no gonio to Q in A-1, autoselect -> Q_lab     OK
No UB, no gonio to Q in A-1, Q_lab      -> Q_lab     OK
No UB, no gonio to Q in A-1, Q_sample   -> Nothing   error message is wrong. Don't need UB. What is target frame in the error message?
No UB, no gonio to Q in A-1, HKL        -> Nothing   OK

No UB, w  gonio to Q in A-1, autoselect -> Q_sam     OK  it rotates
No UB, w  gonio to Q in A-1, Q_lab      -> Q_sam   Label wrong it does not rotate, so it should be Q_lab
No UB, w  gonio to Q in A-1, Q_sample   -> Nothing   ?????? it should do what autoselect does. It asks for UB
No UB, w  gonio to Q in A-1, HKL        -> Nothing   OK


W UB,  w  gonio to Q in A-1, autoselect -> HKL     OK  it rotates, does not rescale (as expected it uses only U)
W UB,  w  gonio to Q in A-1, Q_lab      -> HKL   Label wrong it does not rotate, so it should be Q_lab
W UB,  w  gonio to Q in A-1, Q_sample   -> HKL   Label wrong it does rotate, so it should be Q_sample
W UB,  w  gonio to Q in A-1, HKL        -> HKL   OK, same as autoselect

No UB, no gonio to Q in HKL, autoselect -> Q_lab     OK
No UB, no gonio to Q in HKL, Q_lab      -> Q_lab     OK
No UB, no gonio to Q in HKL, Q_sample   -> Nothing   error message is wrong. Don't need UB. What is target frame in the error message?
No UB, no gonio to Q in HKL, HKL        -> Nothing   OK

No UB, w  gonio to Q in HKL, autoselect -> Q_sam     OK  it rotates
No UB, w  gonio to Q in HKL, Q_lab      -> Q_sam   Label wrong it does not rotate, so it should be Q_lab
No UB, w  gonio to Q in HKL, Q_sample   -> Nothing   ?????? it should do what autoselect does. It asks for UB
No UB, w  gonio to Q in HKL, HKL        -> Nothing   OK


W UB,  w  gonio to Q in HKL, autoselect -> HKL     OK  it rotates, does rescale (as expected)
W UB,  w  gonio to Q in HKL, Q_lab      -> HKL   Label wrong it does not rotate, so it should be Q_lab
W UB,  w  gonio to Q in HKL, Q_sample   -> HKL   Label wrong it does rotate, so it should be Q_sample
W UB,  w  gonio to Q in HKL, HKL        -> HKL   OK, same as autoselect

ConvertToMD is very confusing now, with all the options. Here is what the GUI should look like:

• Keep Q dimensions
• If CopyToMD the next selections should be greyed out
• If modQ or Q3d, allow deAnalysis mode
• If Q3D, allow frames. Change the names to something shorter. Try to write a script with something simpler than Q3DFrames='Q (lab frame)'. Spaces and parentheses in a string are terrible
• If Q_lab is selected, the label should reflect it
• If a goniometer is set, Q_sample should work. label should reflect the choice!!!
• If HKL is selected, only then allow units: rlu or A^-1.
• Orthogonal HKL and Q in lattice units should go away

[Alex Buts]

I agree that ConvetToMD is now confusing and think this ticket needs rethinking. Convert to MD in Q3D case needs rotation and scaling matrix. This is what it needs and uses.

To build this matrix is completely another matter and should be implemented through another algorithm (may be the python one). +it would be nice to be able to change default axis labels and adding this possibility to current convertToMD interface is just unfeasable.

[Andrei Savici]

I think you misunderstood my comment. You have only three matrices in the problem: Goniometer, U, and B. If no goniometer, you can get only Q_lab. If Goniometer is set, you will get Q_sample. To get Q_lab, use goniometer identity. If you have UB, you can get QHKL in A^-1 by using only U, and B is identity, you can get QHKL in rlu using UB. You can still get Q_sample, by setting U, and B to identity, and Q_lab by setting U, B, and Goniometer to identity matrices.

For simplicity, the transformation algorithm to Q3D will always use all 3 matrices, but you set Goniometer, U, and B to unity when needed (or not available). And Depending on what your settings are, you can choose the label as well.

[Alex Buts]

refs #6855 fixed sample frame error.

Changeset: fdf49eae3ba2bd1021ef7b6136872c1bd0ea7356

[Alex Buts]

refs #6855 Changed names of the frames visible in scripts&GUI

also added parameter to test if the goniometer was defined.

Changeset: d7e9a3259acf0db6bc4703c7aadc86f73c21f783

[Alex Buts]

refs #6855 Test for new Goniometer property and use it in MDWSDescript

Changeset: 36f822aa0c6f1ce5603e26016c72162f380685d7

[Alex Buts]

refs #6855 this should fix the axis issues.

Changeset: d28d5bf96cbe631710c40d2fc89f04321fd1e4fc

[Alex Buts]

refs #6855 fixing tests

Changeset: 29fc7c8ac9a68eb11b0d0be91d7ea4f8e7cf920b

[Alex Buts]

The frames, rotations and axis should be correct now.

The GUI is still confusing but incorrect vales in e.g. hkl scaling should be ignored and lab/sample frame selected when requested/(information missed)

The GUI changes should be addressed by separate ticket #6958

[Alex Buts]

refs #6855 changed the target frame in the test/ConvertToMDTest.h

I wander, where this special coordinate system has been set-u. Can not find any :(

Changeset: 16d7d6d3bcd04ac019ecf3b7b45c635485415711

[Alex Buts]

refs #6855 changed the target frame in the test/ConvertToMDTest.h

I wander, where this special coordinate system has been set-u. Can not find any :(

Changeset: 16d7d6d3bcd04ac019ecf3b7b45c635485415711

[Alex Buts]

refs #6855 trivial Doxygen fixed.

Changeset: cb8beaaeecbd541007a7e698d653e9cfbf743084

[Alex Buts]

refs #6855 trivial Doxygen fixed.

Changeset: cb8beaaeecbd541007a7e698d653e9cfbf743084

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 fixed sample frame error.

Changeset: fdf49eae3ba2bd1021ef7b6136872c1bd0ea7356

[Alex Buts]

refs #6855 Changed names of the frames visible in scripts&GUI

also added parameter to test if the goniometer was defined.

Changeset: d7e9a3259acf0db6bc4703c7aadc86f73c21f783

[Alex Buts]

refs #6855 Test for new Goniometer property and use it in MDWSDescript

Changeset: 36f822aa0c6f1ce5603e26016c72162f380685d7

[Alex Buts]

refs #6855 this should fix the axis issues.

Changeset: d28d5bf96cbe631710c40d2fc89f04321fd1e4fc

[Alex Buts]

refs #6855 fixing tests

Changeset: 29fc7c8ac9a68eb11b0d0be91d7ea4f8e7cf920b

[Alex Buts]

refs #6855 changed the target frame in the test/ConvertToMDTest.h

I wander, where this special coordinate system has been set-u. Can not find any :(

Changeset: 16d7d6d3bcd04ac019ecf3b7b45c635485415711

[Alex Buts]

refs #6855 trivial Doxygen fixed.

Changeset: cb8beaaeecbd541007a7e698d653e9cfbf743084

[Alex Buts]

refs #6855 This should fix it

exposing Q-frame property to the algorithm interface

Changeset: c6b94bbc178b3c6cf511d2ecf2c8d6f7b1f06612

[Alex Buts]

refs #6855 fixing weird merge artifacts in ConvertToMDComponentsTest

Changeset: 4a0728da6b8849de243df324960a8f8b61c7b800

[Alex Buts]

refs #6855 weird merge artifacts in ConvertToMD.h

Changeset: 7959746feb904fe90890cd1ba6a36bc1701da211

[Alex Buts]

refs #6855 reverting erroneous merge results

Changeset: 00c8105cf042f29fee3e2f4c242c40c4e45570bd

[Alex Buts]

refs #6855 reverted branch to working state

with possible conflicts from W_matrix ticket remaining unsolved

Changeset: e6a93242289225027953b13db0e9618cc8482773

[Alex Buts]

refs #6855 resolving conflicts from different commits.

Merge branch 'feature/6855_ExplicitQFrame' of ​https://github.com/mantidproject/mantid

Conflicts:

Code/Mantid/Framework/MDAlgorithms/inc/MantidMDAlgorithms/ConvertToMD.h Code/Mantid/Framework/MDAlgorithms/test/ConvertToMDComponentsTest.h

Changeset: 0249d30621917885e6aa7c2d0df9e76c843d6f11

[Alex Buts]

refs #6855 fixed sample frame error.

Changeset: fdf49eae3ba2bd1021ef7b6136872c1bd0ea7356

[Alex Buts]

refs #6855 Changed names of the frames visible in scripts&GUI

also added parameter to test if the goniometer was defined.

Changeset: d7e9a3259acf0db6bc4703c7aadc86f73c21f783

[Alex Buts]

refs #6855 Test for new Goniometer property and use it in MDWSDescript

Changeset: 36f822aa0c6f1ce5603e26016c72162f380685d7

[Alex Buts]

refs #6855 this should fix the axis issues.

Changeset: d28d5bf96cbe631710c40d2fc89f04321fd1e4fc

[Alex Buts]

refs #6855 fixing tests

Changeset: 29fc7c8ac9a68eb11b0d0be91d7ea4f8e7cf920b

[Alex Buts]

refs #6855 changed the target frame in the test/ConvertToMDTest.h

I wander, where this special coordinate system has been set-u. Can not find any :(

Changeset: 16d7d6d3bcd04ac019ecf3b7b45c635485415711

[Alex Buts]

refs #6855 trivial Doxygen fixed.

Changeset: cb8beaaeecbd541007a7e698d653e9cfbf743084

[Michael Reuter]

Merge branch 'feature/6849_MD_metadata' into develop into 6856_ConvertToDiffractionMDWS_v2

[Russell Taylor]

Merge branch 'feature/6804_warnings' into develop into 6856_ConvertToDiffractionMDWS_v2

[Michael Reuter]

Merge branch 'feature/6849_MD_metadata' into develop into 6856_ConvertToDiffractionMDWS_v2

[Russell Taylor]

Merge branch 'feature/6804_warnings' into develop into 6856_ConvertToDiffractionMDWS_v2

[Stuart Campbell]

This ticket has been transferred to github issue ​7701