Gravity correction

[Nick Draper]

From RKH:

Due to the wider range of wavelengths (poss 0.5 to 14 angstrom) and longer distances on SANS2d than on LOQ we really do need to put gravitational droop in. The drops are frequently much more than the precision with which we can determine the beam centre, and significant for the Q value calculation at the longest wavelengths. I think it should go fairly easily into the wavelength to Q calculations. Would help to have a global variable to turn gravity on or off.

Drop ~ 0.000313 lambda^{2 L}2 with lambda in angstrom and sample to detector L in metres. (see attached file)

We won’t worry about the very fine detail of the parabolic paths between sample and detector being slightly longer than the straight line one.

There is an issue as to what wavelength the vertical coordinates on the detector correspond to. We could say they are correct at zero angstrom (as per a tape measure survey) or at some given wavelength, say 4 angstrom, which makes it easier to relate to displays of data.

The latter would suit SANS but I suspect the former would be more logical and transferable to other beam lines ?

[Nick Draper]

Moved as part of iteration 18 end

[Nick Draper]

Moved as part of the end of Iteration 19

[Nick Draper]

Required for reflectometers and SANS.

Get Tim, Rob and SANS guys in a meeting room about this one.

[Russell Taylor]

(In [3313]) Add an algorithm Q1D that amalgamates several steps in the SANS data reduction chain and adds the ability to do masking of time bins properly. Still some efficiency improvements that could go in, but it appears to work for now. The SANSReduction script now uses this. Re #755.

[Russell Taylor]

(In [3346]) Fix problem with errors on points being fitted in CalculateTransmission. Re #755.

[Russell Taylor]

(In [3348]) Add option to CalculateTransmission to fit linearly to the transmission curve rather than logarithmically. Re #755.

[Russell Taylor]

(In [3353]) Fix error which led to removal of one bin too many in StripEndZeroes routine. Re #755.

[Russell Taylor]

(In [3356]) Mistake made in handling bin masking in Q1D - had forgotten to reverse the mask list when the data is reversed in the conversion from wavelength to Q. Corrected the calculation of the final error on log fits in CalculateTransmission. Re #755.

[Russell Taylor]

(In [3361]) Index error in Q1D. Re #755.

[Russell Taylor]

(In [3370]) Flag workspaces resulting from CalculateTransmission as distributions - important if they ever get rebinned (which they do, sometimes). Re #755.

[Russell Taylor]

(In [3389]) Fix the way we're rebinning solid angles when there's bin masking going on. Re #755.

[Russell Taylor]

(In [3391]) The same rebinning issue was present in DiffractionFocussing. Re #755.

[Russell Taylor]

(In [3412]) First attempt at correcting for gravity in the SANS data reduction. Re #755

[Russell Taylor]

(In [3413]) Sign error in gravity correction. Re #755

[Nick Draper]

(In [3420]) re #755 updated constant in gravity formula to ( Drop /m) = 3.1336e-07 * (L/m)^{2 * (lambda/Angstrom)}2

[Nick Draper]

(In [3422]) re #755 adjustment to the application of the drop

[Russell Taylor]

(In [3430]) Correct distance calculation to be full sample-pixel distance, not just horizontal component. Re #755

[Russell Taylor]

(In [3591]) Modified the SANS beam centre finding routine to use the Q1D algorithm, and thus allowing correction for gravity. Re #755.

[Russell Taylor]

(In [3716]) Change Q1D to create its output workspace with the input one as parent. Re #755.

[Russell Taylor]

(In [3718]) Update test reference file following changes to Q1D. Re #755.

[Russell Taylor]

(In [3798]) Some tidying up of Qxy and improvements to unit tests. Re #755.

[Russell Taylor]

(In [3800]) Attempt to get test working on Mac. Re #755.

[Russell Taylor]

Assume this is doing the right thing for gravity, as far as its ambition goes (at least I haven't had any complaints). There may be refinements requested later.

[Anders Markvardsen]

Although update Wiki for Q1D a bit to include short description of what the format required for the QutputBinning property and also mention that gravity correction now included.

[Stuart Campbell]

This ticket has been transferred to github issue ​1603