[Indirect] Output of EISF from fitting routines

[Samuel Jackson]

Ticket request from Spencer:

Output of EISF from fitting routines This is defined as the ratio of elastic amplitude to total (elastic plus quasielastic).

ConvFit - this should be trivial In proc confitParsToWS, add extra member Height to the line

if re.search('HWHM$', yAxis) or re.search('Amplitude$', yAxis) or re.search('Height$', yAxis):

this should create an extra histogram in the WS *_Workspace

Quasi - A bit more complicated In proc C2Fw : lines 474-75 increase range to 9; lines 483-490 adds the elastic amplitude; lines 505-9 adds the extra data to WS. Proc eisf, lines 516-532 is new (could be tidied up). Proc QLrun, add lines 350-1 to run eisf. Lines 382-6 saves the WS.

[Samuel Jackson]

Updated version of IndirectBayes

[Samuel Jackson]

Refs #8585 Output height parameter from ConvFit.

Changeset: 415d531717a854da43fe916b17e349ca87e8cff8

[Samuel Jackson]

Refs #8585 Update ConvFit to output Height parameter.

Changeset: 4eb4684ca1c8e8675901871e7f0b02b18d0bedfb

[Samuel Jackson]

Refs #8585 Update C2Fw to support height.

Changeset: 721e122f3082eba59d0fe9f0b4ee296b1b4c5e8c

[Samuel Jackson]

Refs #8585 Update QLines to output height parameter.

Changeset: a49efb3d41c904b1f9b81ddcdd96ee4c808641b9

[Samuel Jackson]

Refs #8585 Fix y axis names and x data.

Changeset: 6d6e254f9be6e22177ec03927739a2f30e8ebf0d

[Samuel Jackson]

Refs #8585 Add more intelligent spectrum selection process to JumpFit

Changeset: 12523c44fcd4c106fc20af5577fac258010c43ba

[Samuel Jackson]

Refs #8585 Remove debugging statement

Changeset: 187c15387310828429edfb8a6a2c3e66f732220a

[Samuel Jackson]

Refs #8585 Fix output of QLines for sequential fit.

Changeset: bc16e9e23fbc2ac87469520caae5421bcce35da3

[Samuel Jackson]

Refs #8585 Update miniplot with correct spectrum.

Changeset: 6f4deff5140b1f71eb0bbd556b46c63c5d8eb28c

[Samuel Jackson]

Refs #8585 Better formatting on combobox

Changeset: 1a79be1bc56b1d1d374aa92c88cbec51409abb08

[Samuel Jackson]

Merge branch 'feature/8585_EISF_output_from_fitting' into develop

Conflicts:

Code/Mantid/scripts/Inelastic/IndirectBayes.py

Refs #8585

Changeset: 1afb3a8c1503ec4b9feff2697e3c406176d29832

[Samuel Jackson]

Refs #8585 Remove default width options from interface

Changeset: acd58784aff47672cc3dcc99cdce6d7cf0e413ff

[Samuel Jackson]

Refs #8585 Add extra check to prevent plotting.

Also remove warning due to integer casting.

Changeset: 8a787f6989642d0804a3586f9806d2285874aa85

[Samuel Jackson]

Refs #8585 Fix compiler warning.

Changeset: 0b144780e108f4d75bd74af50619944fc71e76cb

[Samuel Jackson]

refs #8585 Corrected x data for histograms.

Changeset: 54faf7d470a192627388367694c57ee1dedfe340

[Samuel Jackson]

Refs #8585 Changed plotting functions to work with new spectrum order.

Changeset: f2e736018a299a0173e4804ce2cf141b20e35c4c

[Samuel Jackson]

Refs #8585 Update interface with correct plot options.

Changeset: 2a743425cf0db4761496ea458d43adde3bad13f7

[Samuel Jackson]

Refs #8585 Changed height parameter to only be output once.

Changeset: b08aa94cdcad94f507aa64024d84a3a9cd568ec8

[Samuel Jackson]

Refs #8585 Refactor QLines QL file reading.

Changeset: c22aa5a8b779839fb6193b519e2fe80501662a96

[Samuel Jackson]

Refs #8585 JumpFit should only list the first three widths.

Also swapped index function names from of Quasi output to be consistent with ConvFit and updated plotting indicies.

Changeset: 67986c3939803691c3770e93c3aaa572dbe81b9e

[Samuel Jackson]

To Tester:

The output of both ConvFit in Indirect Data Analysis and Quasi in Indirect Bayes have been modified and will require testing. Additionally, because the output of both of these functions can be passed to JumpFit, it to will also need testing.

Note that this ticket will need to be tested on Windows and built in Release Mode.

Testing ConvFit

• Follow the instructions in the link below and run a sequential fit: http://www.mantidproject.org/Indirect:Indirect_Data_Analysis#Running_ConvFit
• Open the *_Workspace that is created from the fitting
• Check that the additional parameter 'Height' is included in the workspace.

Testing Quasi

• Use an indirect data file and a resolution file for input (e.g. irs26176 and irs26173 from the system tests)
• Ensure the program selected is QL.
• Ensure 'sequential fit' is ticked.
• Run the fitting and open the *_Workspace that is produced.
• Check that an additional height parameter is output for every function (e.g. f1.f0.Height, f2.f0.Height etc.).

Testing JumpFit

Using the *_Workspace produced by running ConvFit/Quasi, supply them as input to the JumpFit tab on the Indirect Bayes interface.

• Check that the width when using the output of Quasi only displays the first 3 width parameters (f1.f1, f2.f1 and f2.f2)
• Using the output from ConvFit should not allow you to select a width.

Additionally:

• Check that the system tests are passing.
• Check that the format of the text axis on the *_Workspace output from ConvFit and Quasi is consistent.
• Code review to check changes are appropriate.

[Samuel Jackson]

(In #8347) It would be best to test and merge #8585 before this testing ticket as changes have been made to both Quasi, ConvFit and JumpFit that could effect this ticket.

[Samuel Jackson]

Refs #8585 Change ConvFit to work with height and output EISF.

Changeset: 010fea1f5c2e626f098cfa64133dbfe9412eb85b

[Samuel Jackson]

Refs #8585 Adding support for EISF output from Quasi.

Changeset: 3b95e4432593770fd25431a6232d2eafbff8d6df

[Samuel Jackson]

Refs #8585 Tidying up.

Changeset: 7ebb0002c821ff8154537a4fb980f051253ba75a

[Samuel Jackson]

Refs #8585 Updated plotting options.

Quasi should now automatically detect which spectra contains the data it needs to plot.

Changeset: e6e59fd9f935c43b9098c2214d07663885f6e35e

[Samuel Jackson]

Refs #8585 Skip attempting to plot if Plot option not set.

Changeset: 971480ff6a9d22dfc7603cee94d80e9303a9f685

[Samuel Jackson]

Refs #8585 Implement correct error calculations.

Changeset: 274b00f7c8abf3ab1df86fd3ae09f729a5f5d494

[Samuel Jackson]

To Tester:

Follow the instructions in comment 22 and additionally:

• Check that an extra entry called EISF is added for every width/amplitude pair in Quasi and ConvFit.
  • This should be: height / (height+amplitude)

[Samuel Jackson]

Merge branch 'feature/8585_EISF_output_from_fitting' into develop

Conflicts:

Code/Mantid/scripts/Inelastic/IndirectBayes.py

Refs #8585

Changeset: 1afb3a8c1503ec4b9feff2697e3c406176d29832

[Samuel Jackson]

Merge branch 'feature/8585_EISF_output_from_fitting' of github.com:mantidproject/mantid into feature/8585_EISF_output_from_fitting

Full changeset: ​a0d7d198db7a529e48322550bc175628b1855fb6

[Samuel Jackson]

Merge branch 'feature/8585_EISF_output_from_fitting' of github.com:mantidproject/mantid into feature/8585_EISF_output_from_fitting

Full changeset: ​31a34c291b4862644025c27a53be303f0dc00846

[Alex Buts]

Merge remote-tracking branch 'origin/feature/8585_EISF_output_from_fitting'

Full changeset: ​82be1a1e0709214863975c1dc024058f89c1428f

[Alex Buts]

All seems work as described. Tests pass and changes to python look good.

Problem identified during the testing are in the ticket #8770.

[Stuart Campbell]

This ticket has been transferred to github issue ​9429