Transfer Iketa-Carpenter values in IRF into GEM_Definition

[Anders Markvardsen]

At the bottom of this ticket see a copy of an IRF file for GEM. In this copy values are given for SIGMA and GAMMA.

I have so far assumed these are related to SigmaSquared and Gamma (as defined in http://www.mantidproject.org/IkedaCarpenterPV) as follows:

SigmaSquared = Sig-0 + d^{2 * Sig-1 + d}4 * Sig-2 Gamma = Gam-0 + d * Gam-1 + d^{2 * Gam-2}

However no longer sure this is necessarily correct. To start with the FullProf TOF-manual explaining peakshape NPROF=13, the IC peakshape function implemented here, that it is a convolution of IC + pseudo-Voigt. However pseudo-Voigt is defined with two parameter: H and a mixing parameter (see FullProf manual page 48), whereas NPROF=13 contain three extra parameters: SigmaSquared, Gamma and Eta!!! So there is an inconsistency here. A question is: "is it possible to estimate an Eta value from the IRF?".

Discuss this with instrument scientists.

An alternative approach to using the IRF, could be to work towards replicating in Mantid the steps needed to create an IRF in the first place.

COPY OF IRF BELOW:

Instrumental resolution function for GEM/ISIS L. Chapon 11/2003 ireso: 5 ! To be used with function NPROF=13 in FullProf (Res=5) ! ---------------------------------------------------- Bank 1 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 1200.0000 1.0000 19000 ! Dtt1 Dtt2 Zero D2TOF 793.406 0.000 0.364 ! TOF-TWOTH of the bank TWOTH 9.39 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 166.868 98.757 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 0.277 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END ! ----------------------------------------------------- Bank 2 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 1500.0000 1.0000 19000. ! Dtt1 Dtt2 Zero D2TOF 1476.065 0.011 1.881 ! TOF-TWOTH of the bank TWOTH 17.98 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 263.595 0.000 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 3.450 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END ! ----------------------------------------------------- Bank 3 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 2000.0000 1.0000 19000 ! Dtt1 Dtt2 Zero D2TOF 2798.554 -0.274 -1.621 ! TOF-TWOTH of the bank TWOTH 34.96 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 287.456 0.000 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 3.645 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END ! ----------------------------------------------------- Bank 3 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 2800.0000 1.0000 19000 ! Dtt1 Dtt2 Zero D2TOF 4869.121 -2.612 -4.127 ! TOF-TWOTH of the bank TWOTH 63.62 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 176.833 0.000 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 4.416 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END ! ----------------------------------------------------- Bank 3 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 3300.0000 1.0000 19000 ! Dtt1 Dtt2 Zero D2TOF 6671.694 -5.778 -5.029 ! TOF-TWOTH of the bank TWOTH 91.30 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 63.410 0.000 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 3.116 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END ! ----------------------------------------------------- Bank 3 ! Type of profile function: Ikeda-Carpenter * pseudo-Voigt NPROF 13 ! Tof-min(us) step Tof-max(us) TOFRG 4500.0000 1.0000 16700 ! Dtt1 Dtt2 Zero D2TOF 9077.306 -11.374 -6.370 ! TOF-TWOTH of the bank TWOTH 154.40 ! Sig-2 Sig-1 Sig-0 SIGMA 0.000 29.321 0.000 ! Gam-2 Gam-1 Gam-0 GAMMA 0.000 0.982 0.000 ! alph0 beta0 alph1 kappa ALFBE 0.734079 32.017204 2.067249 48.734158 END

[Anders Markvardsen]

From detailed discussion with Aziz it was eventually suggested that what appears to be three independent variables sigmaSq, gamma and eta are are implicitely just two, using pseudo-voigt / voigt relation-ship

Have started to add code to this (note pseudo-voigt / voigt relation-ship-equations incorrect in fullprof manual) and good news is that the Iketa-Carpenter fitting may finally behave.

This task is therefore completed and Ticket #1547 will be used to complete the implementation.

[Stuart Campbell]

This ticket has been transferred to github issue ​2393