Muon: dead time from PlotAssymmetryByLog

[Anders Markvardsen]

Mark Telling writes:

Would it be possible to implement dead time corrections to muon data when running the Mantid > Muon > PlotAssymmetryByLog algorithm

In the dialogue box that is launched it would be great if we could have a tick box which allows the user to choose whether to apply the correction or not – for diagnostic purposes

Now that the muon instruments are running at much higher data rates, the option to correction for dead time is becoming imperative

[Anders Markvardsen]

From a further email conversation with Mark:

Yes, I think an option to launch PlotAssymmetryByLog from the Muon analysis interface would be a good thing - again with the option to apply deadtime corrections or not for diagnostic purposed

[Nick Draper]

Moved to backlog at the code freeze for R2.6

[Arturs Bekasovs]

Added new properties to the algorithm.

Refs #7229

Changeset: 3f1e242a3c5d153b49bdb2c64a7df90298208000

[Arturs Bekasovs]

Fixed indentation of the test code.

Refs #7229

Changeset: 4e1085be2438a288cffc04b69f22858b0cc652c1

[Arturs Bekasovs]

Reduced the number of files loaded for tests.

It's was quite slow already, but I am going to add more tests.

Refs #7229

Changeset: 16c0d679085b965111f2febdcd1b46550cae5be1

[Arturs Bekasovs]

Added tests for functionality to be implemented.

Obviously failing for now.

Refs #7229

Changeset: ce6ea4e2a8f268dfc7828be72fae7fbd6fa89ba2

[Arturs Bekasovs]

Fixed the indentation of the ApplyDeadTimeCorr code.

Refs #7229

Changeset: 5950a637d17bee34b31e1ddff8033a14cb130817

[Arturs Bekasovs]

Code clean-up regarding usage of a namespace.

Refs #7229

Changeset: b2ada415fdb816ef41a09859da01454c37bab315

[Arturs Bekasovs]

Methods for applying dead time correction.

Refs #7229

Changeset: 01c59056db7111987c1cd8becf0c28395b269e09

[Arturs Bekasovs]

New algorithm to group detectors using info from Muon Nexus.

Refs #7229

Changeset: 8a41efdb39d1322b19c7bf3e7a649061d79e5140

[Arturs Bekasovs]

Make the algorithm retrieve workspaces from a group, not from ADS.

Otherwise I couldn't use child algorithm to modify workspace before it gets used.

Refs #7229

Changeset: dbc7caaf1063c77c59579dfe412e71275436d0b2

[Arturs Bekasovs]

Use the new algorithm to apply auto-grouping.

It means I can apply dead time correction before grouping happens.

Refs #7229

Changeset: 72c9c1d127f49aa6b75327d507c1ce15948f95b2

[Arturs Bekasovs]

Improve tests a bit.

Refs #7229

Changeset: 1f0d2ac733ebcee54b0814d5466358a6bef77bfb

[Arturs Bekasovs]

Make grouping functions to return output workspaces.

Refs #7229

Changeset: e1e6635c638ae80b6ff546db79f4c030572e4a58

[Arturs Bekasovs]

Code to apply dead time correction.

Is passing the tests.

Refs #7229

Changeset: 509fa76e21d1abc063a7354495a197fec389e366

[Arturs Bekasovs]

Add an include for std::accumulate.

Refs #7229

Changeset: bd9d9e634e2afac7560f772f98d53d0c37883ac5

[Arturs Bekasovs]

Added extensions to test files so they can be found.

Refs #7229

Changeset: 79cd8d6b16a2af4feb95ff827e22baf36b19c7f3

[Arturs Bekasovs]

Tie all the properties to their widgets.

This make the whole thing less error-prone and gives me a list of all the property widgets.

Refs #7229

Changeset: 1df4cacb069ba2faeb4384b88e4290c5833c7ce1

[Arturs Bekasovs]

Unified the code for handling Browse buttons.

Will be easier to add another one now.

Refs #7229

Changeset: 2c0eadf41060fbea94d137f80e8c326f67e316c6

[Arturs Bekasovs]

Added DTC widgets to the UI.

Refs #7229

Changeset: f017052c6c69a163e41970ad2c49c572abd4ad66

[Arturs Bekasovs]

Code for connecting DTC widgets to algorithm properties.

Refs #7229

Changeset: ea460e29dcf350681dffdeff9361224ee2d84790

[Arturs Bekasovs]

Deal with the situation when ws doesn't contain any auto-grouping info.

Plus improve other error messages.

Refs #7229

Changeset: a58edd5ed87c5771fe87a10e6dcb18d4e86e67dd

[Arturs Bekasovs]

Tester:

Changes touch the PlotAssymetryByLogValue used by Muon scientists, so that's the one you will be testing.

As you might see from the commits, I've had to change the way loaded workspaces are used and auto-grouping is applied within the algorithm. Hence first of all, please try to make sure that PlotAsymmetryByLogValue works in the same way as before in case Dead Time Correction is not applied.

1. Read attached calculate_asymmetry.py through and make sure you understand and agree with the way it calculates a numerical asymmetry value for a range of workspaces.
2. Load it to Mantid Python window and make sure applyDeadTimeCorr variable is set to False.
3. Run the script. This should create a workspace "script_result".
4. Now we will try to create the same results using the PlotAsymmetryByLogValue. Open it's dialog.
5. Set the First run to AutoTestData/MUSR00015189.nxs, Last run to AutoTestData/MUSR00015199.nxs and Output workspace to e.g. result. By default, Log value should be set to run_number and Dead Time Correction Type to None, leave other fields untouched.
6. Run the algorithm. Make sure no errors are thrown and it completes.
7. Now compare the plots or the values of the result and script_result workspaces and make sure they are equal. In addition, compare that to the workspace on the screenshot attached, which is the result of running the algorithm in the last stable release, i.e. before the changes.

So knowing that it wasn't broken, we need to check that the Dead Time Correction is applied correctly.

1. Run another script attached, create_dead_time_table.py. It should create a ws dead_times in the ADS with dead times for 64 detectors, all set to some random value.
2. Now, go back to the calculate_asymmetry.py and set applyDeadTimeCorr to True. This will set it to apply Dead Time Correction using dead_times workspace. Run it.
3. Using SaveNexus algorithm save dead_times workspace to some file.
4. Run PlotAsymmetryByLogValue with Dead Time Correction Type set to FromSpecifiedFile, using the file where dead_times workspace was saved. Make sure it runs successfully.
5. Make sure both script_result and result have changed and are equal again.

Now it is only left to check whether FromRunData DTC works.

1. I've attached a MUSR data file, where dead_times stored in a file where changed to be the same as the ones in dead_times workspace. Download this data file.
2. Set both First run and Last run to the file downloaded. Set Dead Time Correction Type to FromSpecifiedFile, set the file to the one you saved dead_times to. Run the algorithm saving the result to workspace result1.
3. Now set Dead Time Correction Type to FromRunData and run the algorithm with Output workspace result2.
4. Compare result1 and result2 and make sure they are the same.

[Arturs Bekasovs]

Result of the PlotAsymmetryByLogValue for MUSR00015189 - MUSR00015199

[Arturs Bekasovs]

Script which calculates asymmetry for a range of workspaces

[Arturs Bekasovs]

Script which creates dead time table with some random data

[Arturs Bekasovs]

MUSR data file with replaced dead times

[Anders Markvardsen]

Merge remote-tracking branch 'origin/feature/7229_muon_plotbylog_deadtime'

Full changeset: ​a77a209359342753ab06ea9bf66128fabae26383

[Anders Markvardsen]

works as described

[Stuart Campbell]

This ticket has been transferred to github issue ​8075