Implement function "neutron back-to-back exponential convoluted with pseudo-voigt"

[Wenduo Zhou]

It is GSAS's neutron function number 3 and Fullprof's number 9 peak profile. It should inherit IPowderDiffractionFunction?.

[Wenduo Zhou]

Added class NeutronBk2BkExpConvPVoigt. Refs #7653.

Changeset: 983784191b4d19487e5094b1b2f07ce6222d3232

[Wenduo Zhou]

Refactor codes. Refs #7653.

Moved some methods belonged to ThermalNeutronBk2BkExpConvPVoigt to IPowderDiffPeakFunction because they are used by NeutronBk2BkExpConvPVoigt, too.

Changeset: b3f4eabf5bd8b5197e9d348574a4556828a63a87

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: 42c72e80b5127dcd2183748b1381e996dce89cfd

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: 4e3ab2ecb13cd080b33cc51c14572fb334305534

[Wenduo Zhou]

Moved some common function to interface. Refs #7653.

Changeset: 79e491630a8039ea4e22856b89a6248263c6325d

[Wenduo Zhou]

Complete coding (almost). Refs #7653.

Changeset: 1fe695fb6b10b6b31da3a9a8704804ff9ddfdaae

[Wenduo Zhou]

Refactor E1() to interface. Refs #7653.

Changeset: 180ecf20970cb8afc7c514dd4f41261702e02e2a

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: 1342c4640898e3296231e662ffbdb27d0aa97db6

[Wenduo Zhou]

Fixed a bug during testing this ticket. Refs #7653.

Changeset: 89777fb61dfc0fd20df953805f74feee165231b9

[Wenduo Zhou]

Added some unit test (incomplete). Refs #7653.

Changeset: ad7a75f0529bebcfd4b8d3f38512353bbe8ef2ea

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: 4b9e26ce1a36a89c39c2d28e96e7f4261d05756c

[Wenduo Zhou]

Added some unit tests. Refs #7653.

Changeset: 340b00a95e755b6fe8f23722c79c6bca653ea9da

[Wenduo Zhou]

Tried to fix windows build. Refs #7653.

Changeset: f496e9e074981ab71c82c02a0524b90badd84335

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: 05bedc78a764c8aec64486ab37136636c621a30b

[Wenduo Zhou]

Checkpointing work. Refs #7653.

Changeset: b5e63499944ff6e2d95641cbcd5386802a9db39b

[Wenduo Zhou]

Checkpointing work (1113). Refs #7653.

Changeset: 830ec7b9df71b931bbf986f8c1a283f07e64d972

[Wenduo Zhou]

Completed unit test. Refs #7653.

Changeset: d3de072dd20394ca0ba00fb667bc617f976c6958

[Anders Markvardsen]

I tried to do the following

1. Open MantidPlot and plot a spectrum

2. activate the fitting. However to my surprised when right clicking on the plot fit window your function does not shown up as peak

At least users at isis, who uses mantidplot for fitting will for this reason not realise that your new function exit and also with not have the graphical tools activated to adjust such a peak.

Looked at code notice IPowderDiffPeakFunction is not inherenting from IPeakFunction, although looking briefly the method required to satisfy this interface seems to be implemented in IPowderDiffPeakFunction

Also as include this peak in the 'Peak' category

[Wenduo Zhou]

IPowderDiffPeakFunction is an interface class for powder diffraction peaks in Le Bail fit and Rietveld refinement. So far, NeutronBk2BkExpConvPVoigt and ThermalNeutronBk2BkExpConvPVoigt are the only 2 types of powder diffraction peaks inheriting this interface. They are used in powder diffraction analysis, but not designed for the purpose to do single peak fitting. Therefore, they do not inherit interface IPeakFunction.

NeutronBk2BkExpConvPVoigt and ThermalNeutronBk2BkExpConvPVoigt are back-to-back exponential peak functions. But the parameters of them are used to construct functions to calculate back-to-back exponential peak function parameters centered at different d-space. Therefore, it is not practical neither to plot the function nor to fit the peak function in MantidPlot's Fit interface.

There was a discussion about this when I implemented ThermalNeutronBk2BkExpConvPVoigt. And the current implementation follows the conclusion of that discussion.

[Wenduo Zhou]

1. Make (H, K, L) as 3 additional peak parameters;
2. Override unfix() such that H, K, L cannot be fitted;

[Wenduo Zhou]

Added HKL to function parameters. Refs #7653.

Changeset: a93c307825f5cb8d1c58a0f765a530d0349c0dd5

[Wenduo Zhou]

How to verify

1. Download attached arg_si_111.dat;
2. In MantidPlot, load this data file;
3. Launch Fit interface on the plot of arg_si_111.dat;
4. Add function 'NeutronBk2BkExpConvPVoigt'. Set the parameters as Height = 1.436480e+06 Dtt1 = 7476.910000 Dtt2 = -1.540000 Zero = -9.227000 Alph0 = 0.0 Alph1 = 0.597100 Beta0 = 0.042210 Beta1 = 0.009460 Sig0 = 1.741260 Sig1 = 5.746910 Sig2 = 0.0 Gam0 = Gam1 = Gam2 = 0.0 Lattice = 5.431360 H = K = L = 1
5. Do evaluation to this function. You can see the model is close to data.

[Vickie Lynch]

Merge remote branch 'origin/feature/7653_add_fp_peak9'

[Stuart Campbell]

This ticket has been transferred to github issue ​8498