Version 9 (modified by smk78, 6 years ago) (diff)

Beta Approximation Work Package

Incorporation of the 'Beta Approximation' or 'Beta Correction' for S(Q) has long been identified as an important, yet often ignored, correction in model fitting. However, implementing it in SasView is non-trivial.

As of Summer 2018 we are starting to lay the groundwork, with the help of two vacation students, one in the US and one in the UK.

They will approach the implementation from both the bottom-up (changing the code kernel to compute amplitudes instead of intensities, etc) and the top-down (the UI, model parameterisation, choice of effective radius, etc).

As part of the implementation process, computations from SasView are also being validated against those from SASfit, FISH and Matlab.

Tasks

  • Understand exactly what the Beta Approximation calculation requires
  • Figure out how the SasView 'middle layer' needs to change to implement the Beta Approximation calculation
  • Decide how best to redesign the UI to allow a User to make use of the Beta Approximation calculation
  • Decide how best to implement the Effective Radius (ER) of interaction in S(Q) calculations (may involve an overhaul of constraints in SasView)
  • Provide unit tests
  • Provide documentation

Guiding References

  • Analysis of small angle neutron scattering spectra from polydisperse interacting colloids. M Kotlarchyk & S‐H Chen. Journal of Chemical Physics 79, 2461 (1983) https://doi.org/10.1063/1.446055
  • A critical examination of the decoupling approximation for small-angle scattering from hard ellipsoids of revolution. DG Greene, DV Ferraro, AM Lenhoff & NJ Wagner. Journal of Applied Crystallography 49, 1734 (2016) https://doi.org/10.1107/S1600576716012929

Contributors

  • Paul Kienzle & Greg Suczewski (student)
  • Richard Heenan & Torin Cooper-Bennun (student)
  • Yun Liu
  • Paul Butler
  • Wojtek Potrzebowski
  • William Heller
  • Steve King

Tickets

Milestone: SasView 5.0.0 (7 matches)

Ticket Resolution Summary Owner Reporter
#1149 qt5 - Separate composition of mixture models in GUI tcbennun tcbennun
#1136 fixed QT5 GUI - rename FitPage0 to Fitpage1 or rename M1 to M0 richardh richardh
#1135 QT GUI - complex constraints (master ticket) tcbennun tcbennun
#1130 fixed QT GUI - cannot fit any S(Q) params if using P(Q)S(Q) model tcbennun
#1119 duplicate external calls to sasmodels needing beta(Q) richardh richardh
#1116 qt5 user level, changes to model tab tcbennun richardh
#1115 wontfix qt5 user level S(Q) tab richardh

Milestone: SasView Next Release +1 (14 matches)

Ticket Resolution Summary Owner Reporter
#1153 Separate composition of mixture models in GUI tcbennun
#1126 Passing additional computed, non-fitting, parameter values back from models richardh richardh
#1122 General documentation for beta(Q) modification of S(Q) richardh richardh
#1117 fixed qt5 user level Data window richardh
#1114 fixed qt5 gui - model parameter table not displaying well in Windows tcbennun richardh
#1113 wontfix qt5 gui - constraints not working? richardh richardh
#1065 Check the parameter names for complex mixture models pkienzle
#1063 Need unit tests for combined models P(Q)S(Q) etc awashington awashington
#1001 allow mixtures with some components being P*S pkienzle
#780 let the user set effective radius in P(Q)*S(Q) pkienzle
#746 wontfix Consider enhanced fitting constraints richardh
#587 Refactor sum/product model pkienzle
#487 Constraints should be relational operator not assignment operators butler
#475 Add volume fraction explicitly in all relevant models smk78

Milestone: sasmodels 1.0 (5 matches)

Ticket Resolution Summary Owner Reporter
#1202 fixed beta approx branch doesn't support ER/VR tests GitHub <noreply@…> pkienzle
#1201 fixed default effective radius type GitHub <noreply@…> pkienzle
#1121 validation of beta(Q) and other calculations against external code richardh
#1120 beta(Q) approx models to edit richardh richardh
#491 adding support for F(Q) in sasmodels wojciech butler

Milestone: sasmodels Next Release +1 (2 matches)

Ticket Resolution Summary Owner Reporter
#1118 middle level - sort out P(Q)S(Q) including beta(Q) richardh richardh
#1096 improve structure factor calculations with beta approximation pkienzle