RichardWaiteSTFC

Would it not be good to standardise where these basis vectors are stored in MDH workspaces - is it not possible for `ConvertWandSCDToQ` to write them to the workspace basis...

> > Would it not be good to standardise where these basis vectors are stored in MDH workspaces - is it not possible for `ConvertWandSCDToQ` to write them to the...

> This calls for some additional integration tests I agree it would be good to cover all that in integration tests - but perhaps not in this PR!

Just tried testing this on IDAaaS (using .tar.xz found here https://builds.mantidproject.org/job/build_packages_from_branch/85/) I got the following error when trying to refine the bank 1 data ``` Line not defined: 52832.78540 0.00000000...

Is this because it works out Q by putting peak on nearest detector? I would still expect d-spacing to be the same as nominal.

Moved the milestone to 6.2 S3 so I don't forget about this but I don't think I'll find time to look into it. @zjmorgan would you have time to look...

> @RichardWaiteSTFC I will look into this That's great thanks! I don't know if it helps but from the look of this code we seem to calculate the d-spacing using...

Also QLab (from which QSample is retrieved) is also bound to the detector direction at least for Peaks https://github.com/mantidproject/mantid/blob/bb269d6b76dd029d3419f5df77cb73da6ccf2496/Framework/DataObjects/src/Peak.cpp#L446 This shouldn't be an issue for LeanPeaks though right?

> @RichardWaiteSTFC That is correct. So the fact that it also affects the LeanPeaks suggests it might have some other cause within PredictPeaks. Agreed, but we still should not be...

Also just found that when you create a peak with QSample (or equivalently HKL) it seems to move it to have QSample of nearest detector... ``` pk = peaks.createPeakQSample([0., 3.14159265,...