ConvertTo MD does not update EI attached to PreprocessedDetectorsWorkspace

[Alex Buts]

As described in the letter from the instrument scientist, preprocessedDetectorWorkspace does not work propertly and this is the the reason for that.

Should be fixed.

=== OK, I think I have found my problem. I am using BuildSQW_FePS3.py to generate sqw files for two different energy runs. From the 75 meV data I am looking at hplots, and from the 32 meV data kplots. The problem comes if I start off looking at one, and making the necessary plots, and then just run the script again for the other energy data set, WITHOUT first deleting the workspace preprDetMER. As long as I remember to delete this workspace in between running the scripts, then it would appear that everything works just fine.

Sorry, it should have occurred to me that this might present a problem, but it could cause difficulties for others in the future, because they, like me, might assume that this workspace would be rewritten each time the script was run.

Best wishes,

Helen

[Alex Buts]

refs #10332 This should fix it.

Changeset: 959eee237ae7693bfb1804b8f4ae53bd3ab05ef2

[Alex Buts]

the script below will run in current Mantid and will produce 3 absolutely identical workspaces. This is wrong.

After the patch applied, it will produce two workspaces (fail on the md2db), and md2c is very different from md2da.

redWS = CreateSimulationWorkspace(Instrument='MAR',BinParams=[-10,0.1,20],UnitX='DeltaE')
AddSampleLog(redWS,LogName='Ei',LogText='12.',LogType='Number');
md2a=ConvertToMD(InputWorkspace=redWS,QDimensions='|Q|',dEAnalysisMode='Direct',MinValues=[0,-10],MaxValues=[8,20],SplitInto='200,200')

try:
	redWS1 = CreateSimulationWorkspace(Instrument='MAR',BinParams=[-10,0.1,20],UnitX='DeltaE')
	#AddSampleLog(redWS,LogName='Ei',LogText='100.',LogType='Number');
	md2b=ConvertToMD(InputWorkspace=redWS1,QDimensions='|Q|',dEAnalysisMode='Direct',MinValues=[0,-10],MaxValues=[8,20],SplitInto='200,200')
except:
	print " Can not produce md2d workspace"


#redWS1 = CreateSimulationWorkspace(Instrument='MAR',BinParams=[-10,0.1,20],UnitX='DeltaE')
AddSampleLog(redWS,LogName='Ei',LogText='100.',LogType='Number');
md2c=ConvertToMD(InputWorkspace=redWS,QDimensions='|Q|',dEAnalysisMode='Direct',MinValues=[0,-10],MaxValues=[8,20],SplitInto='200,200')

[Alex Buts]

refs #10332 Allow incident energy to be defined as eFixed log name

Changeset: f36f3ad3d88b4800e71c3d106e465de44ba47035

[Alex Buts]

refs #10332 Allow different spelling for eFixed

Changeset: c58c25b5803352e62aa8a4ddfcc5584c83604a05

[Alex Buts]

refs #10332 Test failure if Ei is not defined on source workspace

Always fail in direct or indirect mode.

Changeset: 041b9000ccd443fbf0ef459b4edf05896d926b37

[Alex Buts]

refs #10332 Fix min-max which now correctly changed

(the previous values were related to 60mEv incident energy (Shame!)

Changeset: addd9b9a849c34450b60244f37cfd06f030e2fa9

[Alex Buts]

refs #10332 Remove unnecessary check in indirect mode

As indirect instrument would not easily change

Changeset: 85569c2d992b0977898dc266f4b6fe1eac68cbc2

[Alex Buts]

refs #10332 Should fix MergeMD doctest

it was producing incorrect results!!!

Changeset: aa8e31ccde0f65f781e0f683914b89688dc9a1ae

[Alex Buts]

I think, changes to the code are easy to understand as I am updating Ei attached to Preprocessed detectors' workspace from source workspace each time the algorithm is run.

As the result, the calculations fail (see script above) when Ei is not attached to the source workspace and different energy do produce different scattering patterns for the same energy transfer results.

Tester may compare the results with updated energy with the results when preprocessed detectors workspace is not generated at all ("-" for preprocessed detector's workspace name) and verify that these results are identical to the one obtained in the case when incident energy is updated.

[Federico M Pouzols]

All looks fine. The new code checks error conditions and provides informative error messages. I verified that:

• in current Mantid, three workspaces are created when you run the script provided in the 2nd comment: md2a, md2b, md2c, all of them apparently identical

• after changes, md2b is not created, the try/except block shows an error message. md2c is created and is now different from md2a

• If I forget to specify 'Input neutron's energy', I get a ValueError exception with informative message

• different repetitions seems to produce identical output workspaces

• tests and doc-tests pass (the rst expected output has been updated)

[Federico Montesino Pouzols]

Merge remote-tracking branch 'origin/bugfix/10332_updateEi'

Full changeset: ​cb55e79c1ab4e0ff4365c76a5590eab843aa7386

[Stuart Campbell]

This ticket has been transferred to github issue ​11174