Detector stability Diagnositcs algorithm (5)

[Nick Draper]

identify masking list through change between two white beam vanadium runs.

[Steve Williams]

Procedure:

take two WBV run outputs run integrate on each separately divide the two workspaces to get the efficiency variation find the median the values in that output workspace apply a criterion to find problem detectors

The last stages of this are similar to what I implemented in ticket #598 and so I'll re-use a lot of that code. The implementation described above should take 1 day (8th July). Writing the seems more difficult for me, I'd say 2 days for that, does that mean I'm getting it wrong? I will inform Nick when tasks and or deadlines are complete.

[Steve Williams]

(In [2734]) I wrote a small helper class that reduces the amount of code needed to query detectors and access detector maps, don't know how good it is. A new diagnostics algorithm, DetectorEfficiencyVariation, looks at white beam vanadium runs taken before and after measurements and rejects detectors whose performance changes. refs #601

[Martyn Gigg]

(In [2738]) Fix for tests on linux. Re #601

[Steve Williams]

(In [2740]) First attempt to fixing Linux build problems and DoxyGen warnings that don't show in my tests. There are also some corrections to DetectorEfficiencyVariation. refs #601

[Steve Williams]

The excitations group may look at this algorithm and have some more requirements. It currently exists in the main Mantid code as "DetectorEfficiencyVariation"

[Stuart Campbell]

This ticket has been transferred to github issue ​1449