Ticket #11510 (closed: fixed)
Fix precision problem in BraggScattererInCrystalStructure "Position"-property
| Reported by: | Michael Wedel | Owned by: | Michael Wedel |
|---|---|---|---|
| Priority: | major | Milestone: | Release 3.4 |
| Component: | Diffraction | Keywords: | |
| Cc: | Blocked By: | ||
| Blocking: | Tester: | Dan Nixon |
Description
In ticket #11465, precision problems in calculation of equivalent positions in space groups were addressed.
The class Geometry::BraggScattererInCrystalStructure handles atom positions by exposing a PropertyWithValue<V3D>.
During tests with the structure of SiO2 (trigonal, space group 154) and PoldiCreatePeaksFromCell, it turned out that for correctly handling coordinate values of "2/3", this mechanism is not precise enough. No matter how accurate the position is given in the string passed to the algorithm, it is at some point cut off at 6 digits and leads to incorrect equivalent positions.
The property should be converted to a string that is interpreted using the multi-return value feature of muParser. This has the advantage that users can specify coordinates like "1/4", "1/3", "1/2", "2/3" etc. directly and they are evaluated by muParser down to double precision, preventing problems as described above.
On the downside, it will not be possible any more to assign V3D-object directly to the property using setProperty, so the unit tests using this property have to be adjusted.
Change History
comment:2 Changed 5 years ago by Michael Wedel
Refs #11510. Fix unit tests.
Changeset: 079857285dae8a2f3f6d98b3925413fd9f05f6d6
comment:3 Changed 5 years ago by Michael Wedel
Refs #11510. Added system tests
Added system tests for different structures from different crystal systems.
Changeset: 80d479dae59639d14ff83c899ec4a3ad9ea20077
comment:4 Changed 5 years ago by Michael Wedel
Refs #11510. Fixing system test for SiO2
Changeset: ddc9b915f1dab2259e0122a3c8159b3bfbf28a87
comment:5 Changed 5 years ago by Michael Wedel
- Status changed from inprogress to verify
- Resolution set to fixed
This is being verified as pull request #566.
comment:6 Changed 5 years ago by Michael Wedel
Refs #11510. Removing boost dependency for RHEL6
Changeset: 5c6e9d2442ad417901405eb93ffed862e3d39b7e
comment:7 Changed 5 years ago by Michael Wedel
Refs #11510. Using single value expression parsing for win7
Changeset: 1353c52118abfd63d6b089d71ee2a8d810bc2853
comment:8 Changed 5 years ago by Michael Wedel
Refs #11510. Fixing compiler warning on win7
Changeset: a42f85a8d3d2fad3b988a6c86defa627543565e5
comment:9 Changed 5 years ago by Michael Wedel
Refs #11510. Fixing pylint issue in system test
Changeset: 7b8c7aab4788e22db92f4d3ee94bca616d7db9f3
comment:10 Changed 5 years ago by Dan Nixon
- Status changed from verify to verifying
- Tester set to Dan Nixon
comment:11 Changed 5 years ago by Dan Nixon
- Status changed from verifying to closed
Merge pull request #566 from mantidproject/11510_fix_bragg_scatterer_in_crystal_structure_position
Fix BraggScattererInCrystalStructure "Position" property
Full changeset: b611de4a0f6f0829392d136d32f93ba0988c809a
comment:12 Changed 5 years ago by Stuart Campbell
This ticket has been transferred to github issue 12348
Refs #11510. Use muParser to parse atom position strings.
Changeset: a7b91316ecb1731ec5bcf54540b31d473b2ecdeb