public void DistributionMomentTranslations()
{
int n = 4;
foreach (ContinuousDistribution distribution in distributions)
{
if (Double.IsNaN(distribution.Mean))
{
continue;
}
Console.Write(distribution.GetType().Name);
// Convert central moments to raw moments
double[] centralInputs = new double[n];
for (int k = 0; k < n; k++)
{
centralInputs[k] = distribution.CentralMoment(k);
}
double[] rawOutputs = MomentMath.CentralToRaw(distribution.Mean, centralInputs);
Assert.IsTrue(rawOutputs.Length == n);
for (int k = 0; k < n; k++)
{
Assert.IsTrue(TestUtilities.IsNearlyEqual(rawOutputs[k], distribution.RawMoment(k)));
}
// Convert cumulants to central moments
}
}
public void MomentMathConsistency()
{
// We can't be too demanding here, since there can be strong cancelations.
// We take a low number of simple integer cumulants and try to verify consistency.
double[] K0 = new double[] { 0.0, 1.0, 2.0, 3.0, 4.0, 5.0 };
double mu = K0[1];
double[] C1fromK0 = MomentMath.CumulantToCentral(K0);
Assert.IsTrue(C1fromK0[0] == 1.0);
Assert.IsTrue(C1fromK0[1] == 0.0);
Assert.IsTrue(C1fromK0[2] == K0[2]);
double[] M1fromK0 = MomentMath.CumulantToRaw(K0);
Assert.IsTrue(M1fromK0[0] == 1.0);
Assert.IsTrue(M1fromK0[1] == K0[1]);
double[] K2fromC1 = MomentMath.CentralToCumulant(mu, C1fromK0);
Assert.IsTrue(TestUtilities.IsNearlyEqual(K2fromC1, K0));
double[] M2fromC1 = MomentMath.CentralToRaw(mu, C1fromK0);
Assert.IsTrue(TestUtilities.IsNearlyEqual(M2fromC1, M1fromK0));
double[] K2fromM1 = MomentMath.RawToCumulant(M1fromK0);
Assert.IsTrue(TestUtilities.IsNearlyEqual(K2fromM1, K0));
double[] C2fromM1 = MomentMath.RawToCentral(M1fromK0);
Assert.IsTrue(TestUtilities.IsNearlyEqual(C2fromM1, C1fromK0));
}
public void MomentMapTest()
{
Distribution d = new NormalDistribution();
for (int n = 1; n < 11; n++)
{
double[] K = new double[n + 1];
K[0] = 1.0;
if (K.Length > 1)
{
K[1] = 0.0;
}
if (K.Length > 2)
{
K[2] = 1.0;
}
//for (int m = 1; m < K.Length; m++) {
// K[m] = AdvancedIntegerMath.Factorial(m - 1);
//}
double M = MomentMath.RawMomentFromCumulants(K);
Console.WriteLine("{0} {1}", d.Moment(n), M);
}
}
public void MomentMathOrderOne()
{
double mu = 2.0;
double[] M = new double[] { 1.0, mu };
double[] C = MomentMath.RawToCentral(M);
Assert.IsTrue(C[0] == 1.0);
Assert.IsTrue(C[1] == 0.0);
double[] K = MomentMath.RawToCumulant(M);
Assert.IsTrue(K[1] == mu);
M = MomentMath.CentralToRaw(mu, C);
Assert.IsTrue(M[0] == 1.0);
Assert.IsTrue(M[1] == mu);
K = MomentMath.CentralToCumulant(mu, C);
Assert.IsTrue(K[1] == mu);
M = MomentMath.CumulantToRaw(K);
Assert.IsTrue(M[0] == 1.0);
Assert.IsTrue(M[1] == mu);
C = MomentMath.CumulantToCentral(K);
Assert.IsTrue(C[0] == 1.0);
Assert.IsTrue(C[1] == 0.0);
}
public void CumulantToCentralAndRaw()
{
int n = 8;
foreach (UnivariateDistribution d in Distributions)
{
Console.WriteLine(d.GetType().Name);
// Problems with NaN/Infinity
if (d is CauchyDistribution)
{
continue;
}
if (d is StudentDistribution)
{
continue;
}
if (d is FisherDistribution)
{
continue;
}
if (d is ParetoDistribution)
{
continue;
}
// Real problems
if (d is KolmogorovDistribution)
{
continue;
}
if (d is KuiperDistribution)
{
continue;
}
// From cumulants to central and raw moments
double[] inK = new double[n];
for (int r = 0; r < n; r++)
{
inK[r] = d.Cumulant(r);
}
double[] outC = MomentMath.CumulantToCentral(inK);
for (int r = 0; r < n; r++)
{
Console.WriteLine("r={0} K={1} -> C={2} v C={3}", r, inK[r], outC[r], d.CentralMoment(r));
}
for (int r = 0; r < n; r++)
{
Assert.IsTrue(TestUtilities.IsNearlyEqual(outC[r], d.CentralMoment(r)));
}
double[] outM = MomentMath.CumulantToRaw(inK);
for (int r = 0; r < n; r++)
{
Console.WriteLine("r={0} K={1} -> M={2} v M={3}", r, inK[r], outM[r], d.RawMoment(r));
}
for (int r = 0; r < n; r++)
{
Assert.IsTrue(TestUtilities.IsNearlyEqual(outM[r], d.RawMoment(r)));
}
}
}
