iterative PDF to ODF inversion
calcODFiterative solves the PF to ODF inversion problem by iteratively adjusting the kernel width, starting with a uniform ODF.
Iteratively adjusting the kernel width allows to model ODFs from highly irregularly sampled data, as missing information is modelled by assuming volume portions along tori (fibres) justified by a coarser model. This method basically provides an elaborated guess for starting values for PoleFigure.calcODF and may be seen as some form of regularization.
Syntax
odf = calcODFIterative(pf)
odf = calcODFIterative(pf,'halfwidth',5*degree)Input
| pf | PoleFigure |
Output
| odf | reconstructed SO3Fun |
| alpha | scaling factors, calculated during reconstruction |
Options
| kernel | the ansatz functions (default = de la Vallee Poussin) |
| halfwidth | halfwidth of the ansatz functions (default = resolution) |
| resolution | localization grid for the ansatz functions (default = resolution(pf)) |
| thinning | corresponds to zero range, will remove nodes with low volume portion (default = 0.025, i.e. 2.5 of uniform) |
Flags
| nothinning | omit reduction of nodes |
| silent | no output |
References
https://doi.org/10.1007/978-3-658-14941-3_4 Bachmann, F. (2016). Texturbestimmung aus Beugungsbildern. In: Optimierung der Goniometrie zur Texturbestimmung aus Röntgenbeugungsbildern. Springer Spektrum, Wiesbaden.