SANS add option to provide rebinning parameters to event mode data

[Gesner Passos]

If this option is given, the SliceEvent routine will not try to use the binnings from the monitor.

Because the following steps expect a workspace with monitors as the header, the SliceEvent will apply the same binning to the monitor, but must keep the original to use for the normalization to get the benefit of having more time channels.

[Gesner Passos]

(In #8848) The file is in Babylon in Freddie's folder, and it is SANS2D800024183.nxs (It just add the log to the file Freddie created).

To allow the file to go through it was necessary to apply the patch allow_run_simulated_data_reduction.patch.

Two issues were raised here:

1 - #8906 problem in getting the monitors by guessing its path 2 - The need of a definition of the time channels to Rebin the data and get sensible results.

The purpose of having more time channels for the monitor is that when they are to be normalized in wavelenght, the data in the monitors and the data in the detectors will be compatible (wavelenght depends on the TOF and the geometry). Hence, what is important, is to keep the monitors as they are till the conversion to wavelenght.

The process will be the following:

1 - Allow the users to specify the binnings for the event mode data - #8931. 2 - Ensure that NormalizeToMonitor and TransmissionCalc will be working correctly using the Monitors with more time channels.

[Gesner Passos]

In the gui, Reduction Settings tab, between L/R and L/WAV lines suggest we add a

L/EVENTSTIME

with the linear/logarithmic/variable dropdown as used for L/Q

This would only kick in for event mode data. We then have a backwards compatibility issue, running new Mantid on old data will now use the event mode, not the histogram section of the nxs files [CHECK]. So if L/EVENTSTIME is absent from the user file, Mantid has to use the monitor time bins, so ought to now flag a warning if there are say more than 200 bins ???

[Nick Draper]

Bulk move of tickets out of triage (new) to assigned at the introduction of the triage state

[Peter Parker]

Refs #8931 - Allow use of L/EVENTSTIME command.

Changeset: b9bb6e766c083b88f600cff049886f0370e3e862

[Peter Parker]

Refs #8931 - Cover new command with system test.

Changeset: c7f2a86bf6009d28e0b6b208f7913d56976c4432

[Peter Parker]

Refs #8931 - Remove debugging messages.

Changeset: 38834c0934d9d616561910207143c6640678139d

[Peter Parker]

Refs #8931 - Hook up setting to interface.

Changeset: b574c30199f7743d3a2b8fb1e7e7e04943569ed9

[Peter Parker]

Refs #8931 - Add function comment.

Changeset: b61af47a897753060f3fb196b8d1e1b80ec9cb85

[Peter Parker]

This ticket also made changes to the System Tests repo.

Testing the User File Command

• Open up the ISIS SANS interface and select the "SANS2D" instrument.
• Load the "MaskSANS2DReductionGUI_LimitEventsTime.txt" User File in the SystemTests/Data/SANS2D/ folder, which contains the command L/EVENTSTIME 7000.0,500.0,60000.0.
• Go to the "Reduction Settings" tab and make sure the Events field is populated with the correct string.
• Type "22048" in the the Scattering/Sample field back on the first tab, click "Load Data" and then "1D Reduce".
• Rename the 22048rear_1D_1.5_12.5 workspace to "SomeCounts"

Testing the GUI Field

• Go to the "Reduction Settings" tab and replace the contents of the Events field with 3500.0,500.0,80000.0.
• Back on the first tab, click "1D Reduce" again, and rename the resulting 22048rear_1D_1.5_12.5 workspace to "MoreCounts".

Testing an Empty GUI Field

• Go to the "Reduction Settings" tab and clear the contents of the Events field.
• Back on the first tab, click "1D Reduce" again, and rename the resulting 22048rear_1D_1.5_12.5 workspace to "AllCounts".

Plotting "SomeCounts", "MoreCounts" and "AllCounts" in the same graph should show that the command is having an effect in terms of number of counts.

Testing that Instrument Changes Work

• Change instrument to LOQ and make sure the Events field is no longer visible.
• Change back to SANS2D and make sure that the Events field is visible again, but that it is empty.

[Anders Markvardsen]

$ git test start 8931 git fetch -p ERROR: Branch 'feature/8931_add_option_to_provide_rebinning_params_to_event_data ' has not been fully merged to develop, meaning it is not possible to know whether the code is valid on all environments .

Contact Peter Parker on peter.parker@… (or via skype) and ask them to m erge the code to develop using 'git checkbuild' from their branch feature/8931_add_option_to_p rovide_rebinning_params_to_event_data and let you know when it passes testing on the buildservers so you can start testing.

[Peter Parker]

Refs #8931 - Add file to ignore list in LoadLotsOfFiles.

Changeset: 7b1b9a146b9a487b8b1847bc4a960c9e22101a82

[Peter Parker]

Testing as before, but make sure system tests are passing.

[Peter Parker]

Refs #8931 - Missed a comma.

Changeset: 99a78f5db3464c3c584f6e49a7881bd77441deb9

[Anders Markvardsen]

Merge remote-tracking branch 'origin/feature/8931_add_option_to_provide_rebinning_params_to_event_data'

Full changeset: ​c29c118bfcd926fe8312ea08b54e9e26eb5fa887

[Anders Markvardsen]

Merge remote-tracking branch 'origin/feature/8931_add_option_to_provide_rebinning_params_to_event_data'

Full changeset: ​d513e916fa0c9319f869ca84f9a218eca59af00d

[Anders Markvardsen]

Tested that worked for three reductions with different time slicing

[Stuart Campbell]

This ticket has been transferred to github issue ​9774