public class WatsonFitter extends SphericalDistributionFitter
The routines come from "Statistical Analysis of spherical data" by Fisher, Lewis, Embleton (1987).
Constructor and Description |
---|
WatsonFitter() |
Modifier and Type | Method and Description |
---|---|
static double |
fitKappa(double t)
Fits kappa to an eigenvalue.
|
static double |
fitKappa(EigenSystem3D eig)
Fits kappa to a scatter matrix.
|
static double |
fitKappa(EigenSystem3D eig,
Vector3D[] sampleVecs)
Fits kappa for a sample of vectors.
|
static double |
fitKappa(Vector3D[] sampleVecs)
Fits kappa for a sample of vectors.
|
static double |
getBipolarConfidenceCone(EigenSystem3D eig,
Vector3D[] sampleVecs,
double alpha)
Get the semi-vertical angle of the 100(1 - \alpha)% confidence cone around \mu.
|
static double |
ksTest(Vector3D[] sampleVecs,
Vector3D mu,
double kappa)
Calculate goodness according to the Kolmogorov-Smirnov test.
|
static double |
kuiperTest(Vector3D mu,
Vector3D[] sampleVecs) |
static double |
testBipolarRotSymm(EigenSystem3D eig,
Vector3D mu,
Vector3D[] sampleVecs)
Test of rotational symmetry for a bipolar distribution.
|
static double |
testGirdleRotSymm(EigenSystem3D eig,
Vector3D mu,
Vector3D[] sampleVecs)
Test rotational symmetry for a girdle distribution.
|
tBar, tBarEigenSystem, tBarEigenSystem, tBarEigenSystem, tBarEigenSystem
public static double testBipolarRotSymm(EigenSystem3D eig, Vector3D mu, Vector3D[] sampleVecs)
eig
- the eigensystem of the normalized sample scatter matrix.mu
- the mean axissampleVecs
- the sample axes(p. 165)
public static double getBipolarConfidenceCone(EigenSystem3D eig, Vector3D[] sampleVecs, double alpha)
eig
- the eigensystem of the normalized sample scatter matrix.
The mean axis is presumed to be the first eigenvector, so eig should be sorted.sampleVecs
- the sample axes.alpha
- significance == 100(1 - alpha)%. For a 95% cone, alpha == 0.05.(p. 165)
public static double testGirdleRotSymm(EigenSystem3D eig, Vector3D mu, Vector3D[] sampleVecs)
eig
- the eigensystem of the normalized sample scatter matrix.mu
- the polar axis.sampleVecs
- the sample axes.(p. 182)
public static double kuiperTest(Vector3D mu, Vector3D[] sampleVecs)
mu
- the polar / mean axis.sampleVecs
- the sample axes.(p. 56)
public static double fitKappa(Vector3D[] sampleVecs)
(p.176, 189)
public static double fitKappa(EigenSystem3D eig, Vector3D[] sampleVecs)
(p.176, 189)
public static double fitKappa(EigenSystem3D eig)
(p.176, 189)
public static double fitKappa(double t) throws ConvergenceException
t
- should be t1 for a bipolar distribution about e1, or t3 for a girdle distribution about e3.ConvergenceException
(p.176, 189)
public static double ksTest(Vector3D[] sampleVecs, Vector3D mu, double kappa)
(p. 169)