ReflectometryReductionOneAuto

[Owen Arnold]

Max feels quite strongly that we ought to have methods to grab existing IDF parameters and feed them into our new algorithms (#8371 ReduceReflectometryOne). This is the existing way that quick works. See quick and quick_explicit #8520.

A python algorithm would be ideal for this. We probably ought to mirror all the properties of ReflectometryReductionOne, but pick up any defaults that aren't specified from the IDF. This will make it possible to implement the new version of the reflectometry gui later, where options may or may not be provided by the user and they are otherwise expected to come from the IDF.

[Owen Arnold]

refs #8617. New ReflectometryReductionOneAuto algorithm.

Changeset: 01c5a185776c448ce71f92f9e35d9d44c153512a

[Owen Arnold]

refs #8617. ReflectometryReductionOneAuto refined and tested.

New tests using a base class to avoid duplication. Base class ensures that signature and behaviour of facade algorithm is exactly the same as the base one. In addition, additional tests check that the values supplied to the wrapped algorithm upon execution are what we would expect from the IDF.

Changeset: 1f37f939068cee8240735fa76a356193dbf810e3

[Owen Arnold]

Tester: The following python code should produce two plots. The two curves in each plot should exactly match up. This feature is covered by quite extensive unit tests.

from mantid.simpleapi import *
from isis_reflectometry import quick
reload(quick)

class ReflectometryQuick(object):
  
    def runTest(self):
        defaultInstKey = 'default.instrument'
        defaultInstrument = config[defaultInstKey]
        
        config[defaultInstKey] = 'INTER'
        LoadISISNexus(Filename='13463', OutputWorkspace='13463')
        LoadISISNexus(Filename='13464', OutputWorkspace='13464')
        LoadISISNexus(Filename='13460', OutputWorkspace='13460')
    
        transmissionRuns = '13463'
        runNo = '13460'
        incidentAngle = 0.7
        
        # Use the new auto algorithm providing minimal info
        new_IvsQ1, new_IvsLam1, thetaOut1 = ReflectometryReductionOneAuto(InputWorkspace='13460', FirstTransmissionRun='13463')
        
        # Use the explicit algorithm suppling everything!
        new_IvsQ2, new_IvsLam2, thetaOut2 = ReflectometryReductionOne(InputWorkspace='13460', WavelengthMin=1.0, WavelengthMax=17.0, WavelengthStep=0.05, MonitorBackgroundWavelengthMin=15.0, 
        MonitorBackgroundWavelengthMax=17.0, I0MonitorIndex=2, MonitorIntegrationWavelengthMin=4.0, MonitorIntegrationWavelengthMax=10.0, WorkspaceIndexList=[3,4], FirstTransmissionRun='13463')
        
        plotSpectrum([new_IvsLam1, new_IvsLam2], [0])
        plotSpectrum([new_IvsQ1, new_IvsQ2], [0])
        
        
        
test = ReflectometryQuick()
test.runTest()

[Owen Arnold]

Revert "refs #8617. ReflectometryReductionOneAuto refined and tested."

This reverts commit 1f37f939068cee8240735fa76a356193dbf810e3.

Changeset: 92a4205528842ed57a7e03af37c86edd58d451c3

[Owen Arnold]

Revert "Revert "refs #8617. ReflectometryReductionOneAuto refined and tested.""

This reverts commit 92a4205528842ed57a7e03af37c86edd58d451c3.

Changeset: 87bc7a52ac4bd40aaa2d40c761c6ff78fa8a56c9

[Owen Arnold]

Works on all platforms.

[Karl Palmen]

Similar situation to test for #8616.

[Owen Arnold]

There is no problem here (see my comment for #8616). Look closely, there are two identical curves on each plot.

[Karl Palmen]

Test passes on similar technique as applied it #8616.

[Karl Palmen]

Merge remote-tracking branch 'origin/feature/8617_reduction_auto'

Full changeset: ​a97e8a765a1439cb59cfb686be105f3ee0aec355

[Stuart Campbell]

This ticket has been transferred to github issue ​9461