Geometry

Geometry modules translate detector indices and instrument metadata into physical coordinates used by corrections and integration.

PixelCoordinates3D uses explicit or NeXus-derived detector frames to locate pixel centers in laboratory coordinates. XSGeometryFromPixelCoordinates derives scattering quantities such as Q, azimuth, and solid angle from those positions and beam metadata.

AngleToQ converts scattering-angle or Bragg-angle coordinates to signed Q using a configured or measured centre and photon metadata loaded from an IO source. AngleToQ and XSGeometryFromPixelCoordinates share the same photon_source, photon_units_source, and photon_uncertainties_sources interface. Photon energy versus wavelength is inferred from Pint dimensionality, and energy-to-wavelength conversion uses BaseData arithmetic so units and separate uncertainty components propagate. Keeping the sign allows asymmetric analyser wings to remain separate until a pipeline selects one or combines them deliberately.

This is a breaking configuration change for XSGeometryFromPixelCoordinates: its former wavelength_source, wavelength_units_source, and wavelength_uncertainties_sources fields are no longer accepted. See the interface migration guide for the direct replacements. Wavelength-specific fields on other correction modules are unaffected.

Prefer NeXus transformation chains when the input describes them correctly. Explicit basis vectors, pitches, beam centers, or sample positions remain available for non-NeXus data and reviewed overrides.

Detector indices are dimensionless; pixel pitches and coordinates are lengths. Preserve that distinction. Build static detector geometry once when possible and reuse it across measurements, while keeping measurement-dependent sample or beam metadata in its appropriate source.