rietx at a glance#

This page says which data rietx fits, what it calculates, and what each feature has been checked against.

A row without a comparison with another code says so, and gives what it was checked against instead. The validation page has the numbers behind each link.

Data rietx fits#

rietx fits powder patterns measured with X-rays or with constant-wavelength neutrons. Each instrument preset sets the geometry and the radiation:

Preset	Instrument
`Instrument.bragg_brentano`	laboratory X-ray, flat plate
`Instrument.debye_scherrer`	X-ray in a capillary, including synchrotron
`Instrument.flat_plate_transmission`	X-ray, flat plate in transmission
`Instrument.constant_wavelength_neutron`	neutron, one wavelength

A joint fit refines several patterns of one specimen together, such as a synchrotron pattern and a neutron one. The structure is shared, and each pattern keeps its own instrument, wavelength and background. A joint fit shares the crystallite size and scales each pattern’s width in degrees by its wavelength.

refine_sequential fits a series of patterns one after another, each starting from the last answer. About a fit follows one fit through its stages.

Neutron time of flight is not refined. Instrument.tof_neutron_bank builds a bank and the readers open one, and a time-of-flight fit is refused by name. The roadmap lists it among the work fenced for a 2.0 series.

Features and what they were checked against#

Feature	What rietx has	Checked against	Read more
Rietveld, laboratory X-ray	flat-plate, capillary and transmission geometry, with the Kα doublet	GSAS on fluorapatite, a flat plate; the NIST certificates of LaB₆ and corundum	About a fit, calibrating an instrument
Rietveld, synchrotron X-ray	multi-phase fits and a refinable wavelength	TOPAS on LaB6 and cBN, TOPAS and GSAS-II on two phases, XND on silicon	About a fit
Rietveld, constant-wavelength neutron	the same model with neutron scattering lengths	a published refinement, and TOPAS 6 on calculated patterns	About a fit
Joint fit of several patterns	one structure, one instrument per pattern, a refinable wavelength	the published combined refinement of one specimen	Fitting an X-ray and a neutron pattern together, joint refinement
Le Bail and Pawley	intensities refined per reflection, atoms held	rietx’s own Rietveld results, since no other code was compared	About a fit
Peak shape and axial divergence	pseudo-Voigt (TCHZ) and Voigt, with Finger–Cox–Jephcoat axial divergence	TOPAS 6, on calculated patterns with the pseudo-Voigt	About parameters, calibrating an instrument
Anisotropic strain	Stephens’s model, with the allowed terms derived from the space group	TOPAS 6, on calculated patterns	About parameters
Background	Chebyshev, P-spline and declared humps	TOPAS 6, on calculated patterns with Chebyshev and P-spline backgrounds	About parameters, adding a hump
Extinction	Sabine extinction	TOPAS 6, on calculated patterns	Parameter map
Texture	March–Dollase and spherical harmonics	spherical harmonics against GSAS-II	Parameter map
Phase fractions	weight fractions with esds	the weighed compositions of the IUCr round robin, and TOPAS on LaB6 and cBN	Finding each phase’s weight fraction, about correlations and esds
Files from other programs	TOPAS `.inp`, FullProf `.pcr` and GSAS `.EXP` read in with their refine flags, and GSAS-II `.gpx` read in	the free set GSAS states in `FAP.EXP`, rebuilt exactly	Opening another program’s refinement, names in other programs

Magnetic structures are in development. The pattern rietx calculates for a magnetic phase matches TOPAS 6’s on one calculated pattern. A refined moment has no comparison with another code yet. The roadmap lists the work.

What the report adds#

Beside Rwp and the table of refined values, rietx returns a report of where the fit misses the data and what that misfit looks like, and a list of diagnostics. A diagnostic names a parameter that sits on a bound, two parameters the data cannot separate, or a phase the data cannot see. About the report describes the three layers. Each refinement stage is also kept as a node in a history, so a stage can be restored and the plan forked.