source: sasview/src/sas/sasgui/perspectives/pr/media/pr_help.rst @ 7b07fbe

ESS_GUIESS_GUI_DocsESS_GUI_batch_fittingESS_GUI_bumps_abstractionESS_GUI_iss1116ESS_GUI_iss879ESS_GUI_iss959ESS_GUI_openclESS_GUI_orderingESS_GUI_sync_sascalcmagnetic_scattrelease-4.2.2ticket-1009ticket-1094-headlessticket-1242-2d-resolutionticket-1243ticket-1249ticket885unittest-saveload
Last change on this file since 7b07fbe was 1abd19c, checked in by lewis, 7 years ago

Add note about error on background value to docs

  • Property mode set to 100644
File size: 2.9 KB

P(r) Calculation

Description

This tool calculates a real-space distance distribution function, P(r), using the inversion approach (Moore, 1980).

P(r) is set to be equal to an expansion of base functions of the type

Φn(r) = 2rsin((πnr)/(Dmax))

The coefficient of each base function in the expansion is found by performing a least square fit with the following fit function

χ2 = (i(Imeas(Qi) − Ith(Qi))2)/(error2) + Reg_term

where $I_{meas}(Q_i)$ is the measured scattering intensity and $I_{th}(Q_i)$ is the prediction from the Fourier transform of the P(r) expansion.

The $Reg_term$ term is a regularization term set to the second derivative $d^2P(r)/d^2r$ integrated over $r$. It is used to produce a smooth P(r) output.

Using P(r) inversion

The user must enter

  • Number of terms: the number of base functions in the P(r) expansion.
  • Regularization constant: a multiplicative constant to set the size of the regularization term.
  • Maximum distance: the maximum distance between any two points in the system.

P(r) inversion requires that the background be perfectly subtracted. This is often difficult to do well and thus many data sets will include a background. For those cases, the user should check the "Estimate background level" option and the module will do its best to estimate it. If you know the background value for your data, select the "Input manual background level" option. Note that this value will be treated as having 0 error.

The P(r) module is constantly computing in the background what the optimum number of terms should be as well as the optimum regularization constant. These are constantly updated in the buttons next to the entry boxes on the GUI. These are almost always close and unless the user has a good reason to choose differently they should just click on the buttons to accept both. {D_max} must still be set by the user. However, besides looking at the output, the user can click the explore button which will bring up a graph of chi^2 vs Dmax over a range around the current Dmax. The user can change the range and the number of points to explore in that range. They can also choose to plot several other parameters as a function of Dmax including: I0, Rg, Oscillation parameter, background, positive fraction, and 1-sigma positive fraction.

Reference

P.B. Moore J. Appl. Cryst., 13 (1980) 168-175

Note

This help document was last modified by Paul Butler, 05 September, 2016

Note: See TracBrowser for help on using the repository browser.