Lattice parameter based PoldiSpectrumDomainFunction

[Michael Wedel]

One feature in the existing POLDI analysis software is fitting of lattice parameters using the Pawley-approach.

PoldiSpectrumDomainFunction is based on d, so another layer needs to be added to the function which calculates d using Miller indices and the metric tensor (which provides the actual parameters for refinement).

[Michael Wedel]

Refs #9992. Extract interface of PawleyFunction into API

This is currently required so the function can be used in the SINQ module. When the POLDI algorithms are in CurveFitting/Algorithms, this interface can be deleted again.

Changeset: 8173c342d8083cd4c60c6632f81bd0c9edd60317

[Michael Wedel]

Refs #9992. Remove unused method from PoldiSpectrumDomainFunction.

Changeset: f8f9040bc86c397cc48ada13234234c88a0bb33c

[Michael Wedel]

Refs #9992. Added PoldiSpectrumPawleyFunction

Changeset: aa9faf00c6cc25772deb0b4f3b90649b62529227

[Michael Wedel]

Refs #9992. Checkpointing work

Results look promising already, need some fine tuning for input/output etc.

Changeset: a9db21b181bd7d9947753acc6a493f060e204c34

[Michael Wedel]

Refs #9992. Optimizing PawleyFunction.

Changeset: 246f32694cee0c59f8eb0e7c022ec1f2401d2f66

[Michael Wedel]

Refs #9992. Optimizing Jacobian.

Changeset: f9c0af74138dc3e506b142c411bd0fa0fa74eb73

[Michael Wedel]

Refs #9992. Using numerical derivatives.

Changeset: fc1dc5023caaa04404c3558310050aa164ee76b7

[Michael Wedel]

Refs #9992. Local values, poldi 1D function.

Changeset: 36b6dfdfd61752c1cb5ff050347c6e74a6d39cd2

[Michael Wedel]

Refs #9992. Removing debug output.

Changeset: 076b43d2c3513110917085f5c132abfb40331858

[Michael Wedel]

Refs #9992. Cleaning up, comments.

Changeset: 24237a95e65ff08a557396780f1c80ec93aeaa59

[Michael Wedel]

Refs #9992. Fixing output, input validation

Changeset: e6f4d144fb9a7e2fc5cfe4054c3dfc8de32daf69

[Michael Wedel]

Refs #9992. Removing fix-call from decorator hook

Changeset: aaaf66f1ca5b0be544a071bf92df3577379d7c18

[Michael Wedel]

Refs #9992. Adding unit tests.

Changeset: 9f952ca898edcb8a5c4247a29bfeac279e9e64c9

[Michael Wedel]

Refs #9992. Adding more comments.

Changeset: b7ea20a8f5978d8d8c10b38d65d87f33c2ef0ecd

[Michael Wedel]

Refs #9992. Added doctest, updated documentation.

Changeset: f3897f7bbd368ae080234d4580ea5ad415af6da6

[Michael Wedel]

Refs #9992. Added system test.

Changeset: e9fc18b8714b82d7ad6b92b121e64a465f1a1449

[Michael Wedel]

This is being verified as ​pull request #492.

[Michael Wedel]

Refs #9992. Setting log level for Fit to Notice

Changeset: bfe9c3c47aae672b84de35bf7a1156ab5652a5d6

[Michael Wedel]

Refs #9992. Removing obsolete test, extra semi-colon

Changeset: 8ebd1b805c9d8121063ef6d05252417932bc8161

[Michael Wedel]

Refs #9992. Accidentally committed changed parameter file

Changeset: d994d78612f76eeb8f969418ba88de46875dce02

[Michael Wedel]

Jenkins retest this please

[Michael Wedel]

This will conflict a bit with #475(​https://github.com/mantidproject/mantid/pull/475), where I changed the error calculation. I will leave the "InProgress" label on this one until the other branch is on master to resolve the conflict.

[Michael Wedel]

Refs #9992. Remove debug statement from system test.

Changeset: 52ccce34c4c9588bf80af02379e753bc4357bf7a

[Michael Wedel]

The ticket is not blocked by #9782 anymore, because the derivatives problem was solved in a ticket closed a few weeks ago.

[Michael Wedel]

All conflicts with the the changes from the other pull request have been merged and all the tests passed again.

I created a new ticket #11461(​http://trac.mantidproject.org/mantid/ticket/11461) for refactoring some parts of PoldiFitPeaks2D for the next release cycle.

[Michael Wedel]

You're probably right, I think I meant to use getPointerToProperty(), which just returns the internally stored pointer. Thanks for spotting this!

[Wenduo Zhou]

Are you going to make the change in this pull request or another ticket?

[Michael Wedel]

I think it's better to fix it here, as it's a real mistake and the code freeze is coming soon. I'm pushing the update now, including the merge of master without MDEvents.

[Michael Wedel]

Refs #9992. Fixing memory leak

Changeset: 3c85a23091004716427c68799764bfc479b5788c

[Wenduo Zhou]

I am going to close the ticket as soon as all the checks are passed.

[Michael Wedel]

Thanks!

[Wenduo Zhou]

All tests are passed.

[Wenduo Zhou]

"This pull request is closed, but the 9992_add_poldi_pawley_function branch has unmerged commits. " Is there any commit that has not been merted?

[Michael Wedel]

Did you maybe hit the "Close" instead of the "Merge" button accidentally?

[Wenduo Zhou]

I might have made this mistake.

[Wenduo Zhou]

Merge pull request #492 from mantidproject/9992_add_poldi_pawley_function

Extend PoldiFitPeaks2D to include the option for Pawley type fit

Full changeset: ​c0bce8852086cef7d18cef3e9b6784322fdd6352

[Stuart Campbell]

This ticket has been transferred to github issue ​10834