updates to multidimensional histo writer

[granrothge]

This updates the writing of spec to MDH file writer. So now the output along the x axis is scaled to the proj vector along it's direction rather than the norm of that vector this means a cut 0,0,0 to 1,1,0 would read 0 to 1 along the x axis rather than 0 1.41.
also a bug was fixed so it no properly writes things when q0 is something other than 0,0,0.

The tests were updated accordingly to ensure this bug is captured.

[granrothge]

@mducle @RichardWaiteSTFC Hi guys I caught a bug in my MDH writer that did not handle the case where the cut started from something other than 0,0,0. also I improved the axis scaling. let me know if you have any comments.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (950ad3b) 40.51% compared to head (ec22175) 40.52%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #166      +/-   ##
==========================================
+ Coverage   40.51%   40.52%   +0.01%     
==========================================
  Files         240      240              
  Lines       15981    15985       +4     
==========================================
+ Hits         6474     6478       +4     
  Misses       9507     9507              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[mducle]

Code looks ok, except for a comment...

Tested with this script:

spec = sw_egrid(spinwave(sw_model('triAF', 1), {[-3 -3 0], [1 1 0], 2000}));
spec = sw_instrument(spec, 'dE', 0.2);
proj = [[1 1 0]' [1 -1 0]' [0 0 1]'];
dproj = [1, 1e-6, 1e-6];
sw_spec2MDHisto(spec, proj, dproj, 'test.nxs');

And loaded the resulting test.nxs in Mantid 6.8 and it plots fine - the q-direction scale goes from -3 to +1 as expected (rather than -4.24 to +1.41 as it would before this PR).

[mducle]

I'm not sure this is needed since qidx is always the same as idx...

[granrothge]

Yes the fix was so it would go from -3 to 1 as expected. I updated to remove the redundancy in the variable. I chose to keep qidx and remove idx as it is a more descriptive.

[granrothge]

This should be good to go now.