MUON: PlotAsymmetryByLogValue inconsistent spectrum numbering

[Anders-Markvardsen]

James writes:

We were trying to track down a faulty detector that was upsetting our data using PlotAsymmetryByLogValue. If I enter a range of histograms to use as Forward or Backward (such as "1-16" or "1,3,5,7" it uses Spectra IDs (1-64). But if I enter a single number it instead interprets that as a Workspace Index in the range 0-63.

To test this you could look at HiFi runs 53672 to ~53692 where detector 23 (on the Spectrum numbering scale 1-64) was misbehaving (low counts) in 53672, 53675 and 53678. It's clearly number 23 as seen from plotting one Bin out of the raw data, or 2D Colour Plot.

Forward="1-32" Backward="33-64" or blank (default) -> small spikes seen.

Odd numbers only (long list typed in) -> larger spike.

Even numbers only -> clean data.

Forward="1-22,24-32" and Backward="33-64" -> clean data.

Forward="23" Backward="24" -> nothing wrong with the data...?

Forward="22" Backward="23" -> negative spikes, meaning the low counts are in the Forward bank

Forward="21" Backward="22" -> positive spikes, meaning the low counts are in the Backward bank.

(This is with the very latest Mantid 2.5.3, Windows 7)

[Anders-Markvardsen]

This issue was originally trac ticket 7313