Modify ConvertToMDEvents algorithm to produce both orthogonal and non-orthogonal system of coordinates

[Alex Buts]

Needed to modify ConvertToMDEventWS algorithm to allow conversion into Q orthogonal, 'hkl' orthogonal, and non-orthogonal HKL coordinate systems.

Needed: 1) A switch in the algorithm with would control the MD events coordinate system. 2) Renaming u,v into proj_u, proj_v and adding to them proj_w 3) Keep U matrix unmodified 4) Add to MDGeometry the transformation matrix instead (W and whatever else is needed), which would keep information about the topology and scale of transformation from Q to the MDEvents coordinates and allow to go back (and forward)

See the discussion on "Convert to Q/HKL -> mdEvents changes" for the details.

[Alex Buts]

refs #4813 Started changes allowing to add new events to existing ws

all things prepared to implementation of the ticket but nothing substantial have been done yet.

Changeset: 6a99bf49a53d379da3ee056992e1b23230db561c

[Alex Buts]

refs #4813 fixing errors

Changeset: cd5906b0a7e528cc5a924deed61cbd797e46a706

[Alex Buts]

refs #4813 Started changes allowing to add new events to existing ws

all things prepared to implementation of the ticket but nothing substantial have been done yet.

Changeset: 6a99bf49a53d379da3ee056992e1b23230db561c

[Alex Buts]

refs #4813 fixing errors

Changeset: cd5906b0a7e528cc5a924deed61cbd797e46a706

[Alex Buts]

refs #4813 Minor Edit allowing to run FitTest.h alone

Changeset: ed94ca0ea6a1563ffc2e15526a6b9e6e48f5297d

[Alex Buts]

refs #4813 rough interface to add events to existing MD event ws.

Mainly stub which should be coordinated with other parts of MD Geometry.

Also, as agreed during the discussion, renamed QhQkQl to Q3D and made changes which would make similar modifications easier in a future.

Changeset: 22a1630f9da458bc2638c64c8176fce3943046d7

[Alex Buts]

refs #4813 Separating ConvertToMDEvents.h and ConvertToHKL algorithm

Preparing to extract all logic, related to Q coordinates transformation into separate class outside the ConvertToMDEvents algorithm.

Changeset: c06deecf5337343d63b39c9f4f3cc3cd52a27e7f

[Alex Buts]

refs #4813 Minor changes to SliceDescription

Changeset: 90def14773ced4a946e20fba3d2230f683f3e7a4

[Alex Buts]

refs #4813 rough interface to add events to existing MD event ws.

Mainly stub which should be coordinated with other parts of MD Geometry.

Also, as agreed during the discussion, renamed QhQkQl to Q3D and made changes which would make similar modifications easier in a future.

Changeset: 22a1630f9da458bc2638c64c8176fce3943046d7

[Alex Buts]

refs #4813 Separating ConvertToMDEvents.h and ConvertToHKL algorithm

Preparing to extract all logic, related to Q coordinates transformation into separate class outside the ConvertToMDEvents algorithm.

Changeset: c06deecf5337343d63b39c9f4f3cc3cd52a27e7f

[Alex Buts]

refs #4813 Minor changes to SliceDescription

Changeset: 90def14773ced4a946e20fba3d2230f683f3e7a4

[Alex Buts]

refs #4813 completely removed transfMatrix from ConvertToMDEvents.cpp

and placed them into MDEvents/inc/MantidMDEvents/MDWSTransfDescr. Not working at the moment as transf matrix is not defined.

Changeset: 07bfec3e0a322dd2fdf50b55c2165ca6188cfb00

[Alex Buts]

refs #4401 changed get_vector to getVector

refs #4813 further step in separating transformation matrix from ConvertToMDEvents

Changeset: f878850796feee35f7ed07c40dabfc36704e3a87

[Alex Buts]

refs #4813 Extracted Q-transformation into separate class

All previous tests enabled and pass, so previous functionality is mainly there.

Changeset: 0cb8378d5511e66fada66f5574bd04ed8ebdcba3

[Alex Buts]

refs #4813 transformation work

but more tests are needed to be sure, they work correctly.

Changeset: 24ebf04c1e67e6eb0bdac3cb286e70abc094419c

[Alex Buts]

refs #4813 Seems changes work correct

some unit tests has been added

Changeset: c4f64110a0887f2f48a240e008c39408fb20ffeb

[Alex Buts]

refs #4813 ConvertToCrystallographical system of coordinates

introduced. Very crude algorithm (needs improvement), but better than nothing. Correspondent changes to other MD axis related code and tests

Changeset: 77281520b1f363af121513b9d01569321ff5e246

[Alex Buts]

refs #4813 missing file added

Changeset: 8bd91b57fdbf6d0c0f9a0afc4d23595f2fbd7eb7

[Alex Buts]

refs #4813 fixing unix error

Changeset: 8d54b3ddedffe4efaf2cb41473af72aaa2daedc1

[Alex Buts]

This ticket is substantial part of the ticket #4401, so I can not proceed on those without doing some work here.

I have introduced changes, which hopefully are sufficient to satisfy all requests and provide me with interface to proceed further. Main changes are: 1) W transformation matrix has been added to MDGeometry. Contrary to what is written in Mantid document and in classical papers on this subj, this matrix incorporate both rotation and scaling to account for different coordinate systems requested.

2) All stuff dealing with setting getting this matrix moved to separate class in MDEvents (MDWSTransfDescr) in order for firther changes not to affect too much the conversion algorithm body itself.

I hope that the changes are correct have not verified this properly. But this allows me to proceed with #4401 now.

[Alex Buts]

refs #4813 fixing Unix error

Changeset: d64b2ef69062194a7828c114efb96c3995aac61c

[Alex Buts]

refs #4813 fixing unix error

Changeset: 779fd51aefdbf2694a8b1cb5d67d2891dfdd37dc

[Alex Buts]

refs #4813 Proper implementation of ConvertToCrystal V3D method

Changeset: 888291d9367549e1dccbfcfa0d8c6d63a0515c8b

[Alex Buts]

refs #4813 fixing test

Changeset: 497bcda9f1066c7e0d9aa39737bb1604c9caa98c

[Alex Buts]

refs #4813 better transformToCrystal

but still rubbish

Changeset: ace62a9c75722a4271d5f7a3828b8b510b8b9b10

[Alex Buts]

refs #4813 another possibility to calculate Millers

Changeset: 92a661a68f57bb96db480a332962b3518744ea00

[Alex Buts]

refs #4813 Attempt to make correct axis names

in non-rectilinear system. Far from completed and unsure about the results

Changeset: c2d535e81fbb1196646cde4f42df38c211294ed0

[Alex Buts]

refs #4813 fixing unix warning

Changeset: 35f82b7f7b36804cc40403ebff1e307ba636e814

[Andrei Savici]

ConvertToMDEvents changes. Refs #4766. Refs #4813

Some small improvements still outstanding for orthogonal basis names/scaling. In addition to u and v projections, I added w projection axis

Changeset: 41197f908b6c85915a5ccad50e7583ace6c2e222

[Andrei Savici]

ConvertToMDEvents changes. Refs #4766. Refs #4813

Some small improvements still outstanding for orthogonal basis names/scaling. In addition to u and v projections, I added w projection axis

Changeset: 41197f908b6c85915a5ccad50e7583ace6c2e222

[Nick Draper]

Moved at end of release 2.1

[Andrei Savici]

The forced orthogonal axis needs more testing/tweaking, but the rest works

[Michael Reuter]

Sat with Garrett and verified that everything is in the expected locations and looks correct.

[Stuart Campbell]

This ticket has been transferred to github issue ​5660