Issue in ConvertSpectrumAxis / ConvertUnits with Grouped Detectors

[Michael Whitty]

An issue has been noticed through the Elwin tab in IDA in which the result values for Q / QSquared seem to be nonsense (ie, of order 10^{-100) when the spectra have been grouped at the reduction step.}

This must be investigated and remedied. Attached example nexus file showing the problem.

[Michael Whitty]

I think that I've identified the problem as coming from the following lines (108-115) in ConvertSpectrumAxis.cpp

            try {
              std::vector<double> efixedVec = detector->getNumberParameter("Efixed");
              if (! efixedVec.empty() ) 
              {
               efixed = efixedVec.at(0);
               g_log.debug() << "Detector: " << detector->getID() << " EFixed: " << efixed << "\n";
              }
            } catch (std::runtime_error) { /* Throws if a DetectorGroup, use single provided value */ }

when the algorithm encounters a detector group it won't have an Efixed value to use. Possibly try to retrieve a value from the workspace's parameter map.

[Michael Whitty]

(In [7303]) refs #1838 - ConvertSpectrumAxis (indirect), if it can't find an efixed value for a detector will assume it's a detector group and try taking the first value, then if that is unsuccessful will set a value of zero.

[Nick Draper]

Bulk move of tickets to iteration 27, if your ticket is essential for Iteration 26 then move it back.

[Nick Draper]

Sorry I didn't mean to move these ones reverting back to It 26

[Martyn Gigg]

Ran through the steps:

• Running the ConvertToEnergy IRIS file 21360 to produce a file for input to Elwin;
• Input this into the Elwin tab on the Indirect Data Analysis interface;
• Clicked run and then plotted the values from both of the output workspaces

The results seem of a sensible order of magnitude and the curves have a well-defined shape.

[Stuart Campbell]

This ticket has been transferred to github issue ​2685