\(\renewcommand\AA{\unicode{x212B}}\)

Framework Changes¶

Algorithms ¶

New algorithm MagneticFormFactorCorrectionMD v1 to scale the MDEvents by the magnetic form factor.
Algorithm SaveNexusESS now supports append mode, allowing multiple workspaces to be written either one at a time, or as a group workspace, to a single NeXus HDF5 file.
The algorithm EditInstrumentGeometry now adds detectors to a single detector bank, rather than directly to the ComponentInfo root.
A new property AddSpinStateToLog was added to PolarizationCorrectionWildes to give the option to add the final spin state into the sample log of each child workspace in the output group.
On Linux the algorithm profiler is now built by default but to enable profiling, the properties must be set.
A new InputSpinStates and OutputSpinStates property has been added to PolarizationCorrectionFredrikze v1 and PolarizationEfficiencyCor v1 to allow the order of the workspaces in the input and output Workspace Groups to be set.
A new property UngroupDetectors was added to ExtractMask allowing the output MaskWorkspace to expand the spectra to individal detectors.
add new functions for efficiently calculating absolute and relative differences
On Linux the algorithm AlgoTimeRegister (addtime,Intance) is exposed to python to measure time

Fixed a bug in Sample Transmission Calculator (Interfaces > General > Sample Transmission Calculator) interface to restrict entering commas mixed with decimal point in the double spin boxes for Low, Width and High fields
fixes a bug in LoadNexusProcessed when determining the number of workspaces in a NeXus HDF5 file. It now counts the number of root-level “NX_class: NXentry” groups. Previously, it simply counted the number of root-level groups, assuming all were of “NX_class: NXentry”.
Corrects isDistribution property of result of division of two ragged workspaces, now true
fixes bug in CompareWorkspaces that evaluated NaN values as equal to any floating point (including inf and finite values).
adds new flag NaNsEqual to CompareWorkspaces to control how NaN compares to other NaN s.
Fixed a segmentation fault in FindPeaksConvolve algorithm due to a racing condition in the parallel loop. The issue was first observed as a flaky unit test failure in the CI.
The following parameter names in HeliumAnalyserEfficiency v1 have been updated for consistency: GasPressureTimesCellLength is now PxD, GasPressureTimesCellLengthError is now PXDError, StartLambda is now StartX, and EndLambda is now EndX. Any scripts using the old names will need to be updated.

EnumeratedStringProperty which uses EnumeratedString can be used in C++ based algorithms

Fix CrystalStructure to display deuterium when it is one of the atoms
Fixed bug in TableWorkspace::getMemorySize() where the calculation was not summing memory correctly, leading to an underestimate of memory use..

MatrixWorkspace now has a plotType property
Adds TableWorkspaceNotEmptyValidator
allows for declaring the many PropertyWithValue types as output properties from the python API.
Added a new testing function assert_not_equal to make testing inequality between workspaces more convenient.

Drop support for NumPy version 1. We now build against NumPy v2.0 and support up to v2.1. Read about the changes. Users should note that NumPy 2 introduces some breaking API changes. See the `NumPy 2 Migration Guide <https://numpy.org/devdocs/numpy_2_0_migration_guide.html>`_ for more details
Updated compiler on Linux to gcc version 13, which should improve performance in some circumstances. The release notes can be found here https://gcc.gnu.org/gcc-13/changes.html
Updated Matplotlib from version 3.7 to version 3.9. The release notes for version 3.8 and version 3.9