Version 60 (modified by pkienzle, 7 years ago) (diff)

Add a model here when you are working on it. Mark it as converted when it is tested and pushed.

The err columns are the result of running multi_compare.sh, which generates random parameters for comparison against sasview. double should be at most 5e-14 relative error, single should be 5e-5.

./multi_compare.sh all 200 1d200 mono double
./multi_compare.sh all 200 1d200 mono single

Some models still have issues when run using single precision (lamellarPC for example).

We are not dealing with magnetism at this point as it requires some work on the wrapper functionality. Thus the *Magnetic* column below is for reference as to which models need attention. The *SESANS* column should be used to indicate if the model is a priority (1 highest, 4 lowest).

SasView namesasmodels name P(Q) S(Q) Multiply Multiplicity Magnetic SESANS Assigned Converted double err single err
HayterMSAStructure HayterMSAsq NYNNN1RKHY 0, 3/200 NaN FAIL
BarBellModel barbell YNNNN3AJJY good good
BCCrystalModel bcc_paracrystal YNNNN3AJJY good FAIL
BinaryHSModel binary_hard_sphere YYNNN1PDBN
BroadPeakModel broad_peak YNNNN4PDBY good good
CappedCylinderModel capped_cylinder YNNNN3HP Y good good
CoreShellCylinderModelcore_shell_cylinderYNYNN3HP Y good good
CylinderModel cylinder YNYNN - but doc says so1HP Y good good
!DABModel dab YNNNN1DM Y good good
EllipsoidModel ellipsoid YNYNN1HP Y good* good
FCCrystalModel fcc_paracrystal YNNNN3AJJY good FAIL
PeakGaussModel gaussian_peak 4AJJN good 1e-3
!Guinier guinier YNNNN1JRKY good good
HardsphereStructure hardsphere NYNNN1RKHY needs utest good FAIL
HollowCylinderModel hollow_cylinder YNNNN2JRKY good 2e-3
LamellarModel lamellar YNNNN2PAKY good good
LamellarPSModel lamellarCaille YNNNN3RKHY needs utest good 5e-2
LamellarPSHGModel lamellarCailleHG YNNNN3RKHY needs utest 3e-12 5e-3
LamellarFFHGModel lamellarFFHG YNNNN3RKHY needs utest good good
LamellarPCrystallModel lamellarPC YNNNN3RKHY needs utest good FAIL
!Lorentz lorentz YNNNN1JRKY good good
ParallelepipedModel parallelepiped YNYNN - but doc says so2MAGY good good
!AbsolutePower_Law power_law 3SMKN good good
SphereModel sphere YNYNY1HP Y good* good
StickyHSStructure stickyhardsphere NYNNN1RKHY needs utest good FAIL
TeubnerStrey teubner_strey 1PAKY good* good
TriaxialEllipsoidModeltriaxial_ellipsoid YNYNY3HP Y good* good
FuzzySphereModel 2TRNN
RaspBerryModel 2AJJ
CoreShellModel doc says so2MDN
CoreMultiShellModel doc says so3
!Core2ndMomentModel 3SMKN
MultiShellModel 3PAK
OnionExpShellModel 3
VesicleModel 2PDBN
!SphericalSLDModel 3
LinearPearlsModel 3
PearlNecklaceModel 3JRKN
EllipticalCylinderModel 3MD N
FlexibleCylinderModel 2 N
!FlexCylEllipXModel 3
CoreShellBicelleModel 2
StackedDisksModel 2TRNN
PringleModel 3AJJ
CoreShellEllipsoidModel 3
!CoreShellEllipsoidXTModel 3
!SCCrystalModel 3
!CSParallelepipedModel 3MAGN
RectangularPrismModel 2MAGN
RectangularHollowPrismModel 3MAGN
RectangularHollowPrismInfThinWallsModel 3MAGN
!Debye 1SMKN
CorrLength 3
FractalModel 1PDB
MassFractalModel 1
SurfaceFractalModel 1
MassSurfaceFractal 1
FractalCoreShell 3
GaussLorentzGel 2
!BEPolyelectrolyte 3
GuinierPorod 2MD N
PorodModel 3MD N
PeakLorentzModel 4AJJN
!Poly_GaussCoil 2SMKN
PolyExclVolume 3
!RPA10Model 3PAK
TwoLorentzian 4
TwoPowerLaw 4
UnifiedPowerRg 4
LineModel 4
GelFitModel 1
StarPolymer 2
ReflectivityModel 4
!ReflectivityIIModel 4
SquareWellStructure 1PDB
!testmodel 3
!testmodel_2 3
!sum_p1_p2 3
!sum_Ap1_1_Ap2 3
!polynomial5 3
!sph_bessel_jn 3

* using taylor expansion near 0 for fq = 3 (sin(qr)/qr - cos(qr))/qr2; using quad precision as a cross check, this is more accurate than computing the direct expression for qr < 0.1 in double precision, and so the values will differ from those in sasview. It is much more accurate in single precision. This affects many models, including sphere and ellipsoid models. The cylinder models are using J1(qr)/qr, which has its own numerical expansion and does not suffer from this problem. Should probably settle on one or the other and move it into the library. So even though the double precision values for sphere, ellipsoid and triaxial ellipsoid differ by more than 5e-14, they are still marked good. Similarly, the Teubner-Strey model checks out against the quad precision version and so is marked good.

1GList of standard parameters (name = meaning):

  • radius = radius
  • sld = scattering length density of particle
  • solvent_sld = scattering length density of matrix
  • cor_length = correlation length
  • exp = exponent (example: porod_exp)
  • peak_pos = q_peak or q0 etc
  • theta, phi, psi