1D Fit function to optimize an interpolated structure factor

[Jose Borreguero]

A 1D function that serves a interpolated structure factor S(Q,E,f) versus some parameter f. This structure factor is constructed from a discrete set of {S(Q,E,f_i)} structure factors.

The showcase is a series of molecular dynamics simulations where parameter f is one force-field component of interest. Simulations carried out under different values of this parameter yield different structure factors, giving rise to the discrete set {S(Q,E,f_i)}. We want to construct an interpolator with this set and then optimize against experimental data, thus obtaining the value of the force-field component f yielding the simulation closest to experiment.

TESTING THE TICKET: It is first necessary to install python package dsfinterp, which can be obtained from the python package index

sudo pip install dsfinterp

If the package is not installed, then the function will not be subscribed and will be unavailable.

Uncompress & untar attached file, preferrably under '/tmp'. Open MantidPlot, then load and execute script DSFinterp1DFitTest.py. You can read the docstring in the script for a detailed explanation, but you should check that parameters "f1.TargetParameter" and "Cost function value" in workspace "exp200K_Parameters". These parameters should evaluate in the neighbourhood of 0.064115 and 6.90233, respectively.

[Jose Borreguero]

Refs #9081 First pass at the fit function

Changeset: 0dd5b33f28152c4db99d0db7e40430745f4397ff

[Jose Borreguero]

Refs #9081 More modifications to the fit function

Changeset: d993779035b1eb95c5e82f95fe36fe2f37b940ea

[Nick Draper]

Please make sure the algorithm help page specifies the need for the pre-requisite and instructions of how to get it.

[Jose Borreguero]

Refs #9081 More modifications

Changeset: 5b1b0a3d8b500918eeabbacc4d20253cbc1a7395

[Jose Borreguero]

Refs #9081 Fixing some bugs

Changeset: cb86f4c42313c73392a783aa61a356d5c6b92dc9

[Jose Borreguero]

Refs #9081 First pass at the test file

Changeset: 5d3a7eb5686809cba043ee171b4d374876c3a682

[Jose Borreguero]

Refs #9081 Debug test file

Changeset: 7971b5645dc6ffa570130b0cb24bd918aeab1ef8

[Jose Borreguero]

Refs #9081 Small clean up

Changeset: 8f1485931431ca8030bb2ca183bebb5b9f6454bc

[Jose Borreguero]

Refs #9081 Various fixes

Changeset: 597fd8ec8b05a30991400518f81d89ed9bb66f6f

[Jose Borreguero]

Refs #9081 Small modifications

Changeset: d3ce13b5047554a36b7d1c82e185c20a0577c445

[Jose Borreguero]

Refs #9081 comment the import of the python debugger

Changeset: 4a5c5ebce29b5f7c7e6096f0c332c4aca297cd32

[Jose Borreguero]

Refs #9081 Add linear interpolator

A linear interpolator is needed when the energies for the channels do not coincide with the fitting domain

Changeset: 8db47e2cb473855a0b07a127135240ebb0c06607

[Jose Borreguero]

Refs #9081 Added WIKI comments

Changeset: f1484347f9223c33a1ce479a70b82a10985a371f

[Jose Borreguero]

Refs #9081 Remove silly single quotes in the WIKI page

Changeset: eb363c58c8e1770944340c262ec04becbb8c1133

[Jose Borreguero]

TESTING THE TICKET: It is first necessary to install python package dsfinterp, which can be obtained from the python package index

sudo pip install dsfinterp

Uncompress & untar the attached file, preferrably under '/tmp'. Open MantidPlot, then load and execute script DSFinterp1DFitTest.py. You can read the docstring in the script for a detailed explanation, but you should check that parameters "f1.TargetParameter" and "Cost function value" within workspace "exp200K_Parameters". These parameters should evaluate in the neighbourhood of 0.064115 and 6.90233, respectively.

[Vickie Lynch]

Merge remote branch 'origin/feature/9081_DSFinterp1DFit'

Full changeset: ​d5681f5fb69f234a191cac51c4bb33bf28e16f71

[Stuart Campbell]

This ticket has been transferred to github issue ​9924