7.1.1 d-Dimensional Symmetric Elliptical Type Distribution

In previous chapters, we have discussed symmetric and asymmetric (i.e., nested) Archimedean copulas for multivariate modeling (i.e., d ≥ 3d≥3). We have shown that (i) symmetric multivariate Archimedean copulas require that all the correlated variables share the same dependence structure, and (ii) the nested asymmetric multivariate copulas still require that some variables share the same dependence structure. Compared to the symmetric and nested asymmetric Archimedean copulas, the meta-elliptical copulas are more flexible than the symmetric (or nested) Archimedean copulas for modeling multivariate hydrological variables.

Following Genest et al. (2007), a d-dimensional random vector z (z = [z₁, …, z_d]^Tz=z1…zdT) is said to have an elliptical joint distribution, i.e., ℰ_d(μ, Σ, g)ℰdμΣg with mean vector μ(d × 1)μd×1, covariance matrix Σ(d × d)Σd×d, and generator g : [0, ∞)→[0, ∞)g:0∞→0∞, if there exists a stochastic representation, as follows:

where r ≥ 0r≥0 is a random variable with the probability density function as

fgr=2πd2Γd2rd−1gr2(7.1a)

u (independent of r) is uniformly distributed on the sphere as follows:

Sd=u1…ud∈ℝd:u12+u22+…+ud2=1(7.1b)

A is Cholesky decomposition of Σ, i.e., AA^T = ΣAAT=Σ

and the joint probability density function of z can be written as follows:

Σ−12gz−μTΣ−1z−μ(7.2)

In Equations (7.1a) and (7.2), g(⋅)g⋅ is a scale function uniquely determined by the distribution of r and referred to as the probability density function generator. Common d-dimensional symmetric elliptical type distribution generators are given in Table 7.1.

Table 7.1. Common probability density function generators [g(t)gt] for elliptical copulas.

Copula	g(t)gt
Normal	2π−d2exp−t2
Student	πv−d2Γd+v2Γd21+tv−d+v2
Cauchy	π−d2Γd+12Γ121+t−d+12
Kotza	sΓd2r2N+d−22stN−1exp−rtsπd2Γ2N+d−22s;r,s>0,2N+d>2
Pearson type II	Γd2+m+1πd2Γm+11−tm;t∈−11,m>−1
Pearson type VIIb	ΓNΓN−d2πmd21+tm−N;N>d2,m>0

Notes:

^a Kotz type copula reduces to normal copula if N = s = 1, r = 1/2N=s=1,r=1/2.

^b Pearson type VII copula reduces to Cauchy copula if m = 1, N = 3/2m=1,N=3/2 and reduces to Student copula if N=m2+1.

To build the meta-elliptical copula using the g(⋅)g⋅ function (listed in Table 7.1) and Equation (7.2), we should note that there is one limitation of these elliptical distributions, that is, the scaled variables z1σ11,z2σ22,…,zdσdd are identically distributed with the density function as follows:

qgzkσkk=x=πΓd−12∫u2∞y−x2d−12−1gydy;k=1,…,d(7.3)

and the CDF of the scaled variables given as follows:

Qgzkσkk≤x=12+πd−12Γd−12∫0x∫u2∞y−u2d−12−1gydydu(7.4)

From Equations (7.3) and (7.4), it is known that q_g(x) = q_g(−x)qgx=qg−x and Q_g(x) = 1 − Q_g(x) forx>0Qgx=1−Qgxforx>0.

Example 7.1 Derive the d-dimensional multivariate normal density function: z = [z₁, …, z_d]z=z1…zd.

Solution: As listed in Table 7.1, the probability density function generator for the multivariate normal distribution is gt=2π−d2exp−t2. Applying Equation (7.2), we have the following:

fz=Σ−122π−d2exp−z−μTΣ−1z−μ2,z~ℰdμΣg(7.5)

If μ = 0μ=0 in Equation (7.5), we have

fz=Σ−122π−d2exp−zTΣ−1z2,z~ℰd0Σg(7.6a)

By applying Equation (7.1), we have z^TΣ⁻¹z = r²(Au)^TΣ⁻¹(Au) = r²zTΣ−1z=r2AuTΣ−1Au=r2 and Equation (7.6a) may be rewritten as follows:

fz=Σ−122π−d2exp−r22,z~ℰd0Σg(7.6b)

where

Σ=ρ11⋯ρ1d⋮⋱⋮ρd1⋯ρdd,ρii=1;∣ρij∣<1,i≠j;i,j=1,..,d, correlation matrix.

Example 7.2 Derive the d-dimensional multivariate Cauchy density function for z = [z₁, …, z_d]z=z1…zd.

Solution: Using the probability density function generator for the multivariate Cauchy distribution listed in Table 7.1:

gt=π−d2Γd+12Γ121+t−d+12

Applying Equation (7.2), we have the following:

fz=Σ−122−d2π−dΓd+12Γ121+z−μTΣ−1z−μ2−d+12,z~ℰdμΣg(7.7)

Similarly, if μ = 0μ=0, we have from Equation (7.7) the following:

fz=Σ−122−d2π−dΓd+12Γ121+zTΣ−1z2−d+12,z~ℰd0Σg(7.7a)

Or equivalently

fz=Σ−122−d2π−dΓd+12Γ121+r22−d+12,z~ℰd0Σg(7.7b)

Without loss of generality, we will only investigate the case ℰ_d(0, Σ, g)ℰd0Σg. Let z = [z₁, z₂, …, z_d]^Tz=z1z2…zdT be a random vector with each component z_izi with given continuous PDF f_i(z_i)fizi and CDF F_i(z_i)Fizi. Suppose

xi=Qg−1Fizi,i=1,2,…d(7.8)

where Qg−1 is the inverse of Q_gQg.

Then, the probability density function of z is given by

where the Jacobian matrix J is given as follows:

J=∂x1∂z1⋯∂xd∂zd⋮⋱⋮∂x1∂zd⋯∂xd∂zd

Since xi=Qg−1Fizi, we have ∂xi∂zj=dxidzi,i=j0,i≠j. Rewriting matrix J, we have the following:

J=dx1dz1⋯0⋮⋱⋮0⋯dxddzd;J=dx1dz1⋅dx2dz2⋯dxddzd=∏i=1ddxidzi

From xi=Qg−1Fizi, we have F_i(z_i) = Q_g(x_i)Fizi=Qgxi. Differentiation on both sides leads to f_i(z_i)dz_i = q_g(x_i)dx_ifizidzi=qgxidxi; dxidzi=fiziqgxi=fiziqg(Qg−1Fizi⇒J=∏i=1dfiziqgQg−1Fizi

Then, we have the following:

fz1z2…zd=fx1x2…xd∏i=1dfiziqg[Qg−1(Fizi]=fx1…xd∏i=1dqgxi∏i=1dfizi(7.10)

For x = (x₁, …, x_d)^T~ℰ_d(0, Σ, g)x=x1…xdT~ℰd0Σg, we have the following:

fx1…xd=Σ−12gxTΣ−1x(7.11)

Inserting Equation (7.11) in Equation (7.10), we have the following:

fz1…zd=Σ−12gxTΣ−1x∏i=1dqgxi∏i=1dfizi(7.12)

Using ℋℋ to represent the d-variant probability density function as

ℋQg−1F1z1…Qg−1Fdzd=Σ−12gxTΣ−1x∏i=1dqgxi

Equation (7.12) may be written as follows:

fz1…zd=ℋQg−1F1z1…Qg−1Fdzd∏i=1dfizi(7.12a)

To this end, the d-dimensional random vector z is said to have a meta-elliptical distribution, if its probability density function is given by Equation (7.12). Denote x~ℳℰ_d(0, Σ, g; F₁, …, F_d)x~ℳℰd0ΣgF1…Fd. The function ℋQg−1F1z1…Qg−1Fdzd is referred to as the probability density function weighting function. The class of meta-elliptical distributions includes various distributions, such as elliptically contoured distributions, the meta-Gaussian distributions, and various asymmetric distributions. The marginal distributions Fi. can be arbitrarily chosen (Fang et al., 2002). The meta-elliptical distributions allow for the possibility of capturing tail dependence (Joe, 1997), which will be discussed later.

7.1.2 Bivariate Symmetric Elliptical Type Distribution

Suppose x~ℳℰ₂(0, Σ, g)x~ℳℰ20Σg, we have the following:

Σ=1ρρ1,Σ−1=11−ρ2−ρ1−ρ2−ρ1−ρ211−ρ2

x1x2Σ−1x1x2=x12+x22−2ρx1x21−ρ2

Vector x has the following probability density function:

fx1x2Σg=11−ρ2gx12+x22−2ρx1x21−ρ2(7.13)

The marginal PDF and CDF of x are

qgx=∫x2∞y−x2−12gydy(7.14)

Qgx=12+∫x2∞arcsinxydy(7.15)

A two-dimensional random vector (x₁, x₂)x1x2 follows an elliptically contoured distribution, if its joint PDF takes on the form of Equation (7.13). Its copula function can be written as follows:

CXuv=Fx1x2=11−ρ2∫−∞Qg−1u∫−∞Qg−1vgs2+t2−2ρst1−ρ2dsdt(7.16)

where u=F1x1,v=F2x2,s=Qg−1u,t=Qg−1v.

The copula density function can be given as follows:

where

ℋstρ=fstρqgsqgt(7.18)

Now, let z = (z₁, z₂)^T~ℳℰ₂(0, Σ, g; F₁, F₂)z=z1z2T~ℳℰ20ΣgF1F2. Its probability density function may then be expressed as follows:

fx1x2=ℋQg−1F1x1Qg−1F2x2f1x1f2x2(7.19)

Take simple examples to illustrate the preceding.

Symmetric Kotz Type Distribution

Let x be distributed according to a bivariate symmetric Kotz type distribution. Inserting the density generator for Kotz type distribution (listed in Table 7.1) in Equation (7.2), we obtain the joint probability density function as follows:

fx1x2=srNsx12+x22−2ρx1x2N−1πΓNs1−ρ2N−12exp−rx12+x22−2ρx1x21−ρ2s(7.20)

where r>0, s>0, N>0r>0,s>0,N>0 are the parameters.

The marginal PDF (i.e., q₁(x)q1x) can be written as follows:

q1x=2srNsπΓNs∫0∞t2+x2N−1exp−rt2+x2sdt(7.21)

The corresponding CDF (i.e., Q₁(x)Q1x) can be written as follows:

Q1x=12+2srNsπΓNs∫0x∫0∞t2+x2N−1exp−rt2+x2sdtdx(7.22)

Then, the copula probability density function can be given as follows:

cuv=fQ1−1uQ1−1vq1Q1−1uq2Q2−1v(7.23)

Example 7.3 Show that the bivariate Kotz type distribution converges to the bivariate Gaussian distribution as noted in Table 7.1, i.e., N = s = 1, r = 1/2N=s=1,r=1/2.

Solution: Substituting N=s=1,r=12 into the probability density function generator of symmetric Kotz type distribution, we have

g2t=sΓd2r2N+d−22stN−1exp−rtsπd2Γ2N+d−22s=texp−t/22π,d=2

Comparing with the probability density function generator for the normal copula in the bivariate case, we have the following:

g2N=2π−1exp−t2

Now we show that the bivariate Kotz type distribution reduces to the bivariate normal distribution if N=s=1,andr=12. The same conclusion is reached for higher dimensional cases.

Example 7.4 Compute the copula density function for symmetric Kotz type distribution with the information given as N = 2.0, s = 1.0, r = 0.5, ρ = 0.1, u = 0.4, v = 0.3N=2.0,s=1.0,r=0.5,ρ=0.1,u=0.4,v=0.3.

Solution: Using Equation (7.22), we can calculate Q⁻¹(u), Q⁻¹(v)Q−1u,Q−1v numerically as follows:

Q⁻¹(0.4) =  − 0.4843; Q⁻¹(0.3) =  − 0.9158

Q−10.4=−0.4843;Q−10.3=−0.9158

Using Equation (7.20), we can compute the joint density function as follows:

f(Q⁻¹(0.4), Q⁻¹(0.3)) = f(−0.4843, −0.9158) = 0.0484

fQ−10.4Q−10.3=f−0.4843−0.9158=0.0484

Using Equation (7.21), we can compute the univariate density as follows:

q(Q⁻¹(0.4)) = q(−0.4843) = 0.2190; q(Q(v)) = q(−0.9158) = 0.2411

qQ−10.4=q−0.4843=0.2190;qQv=q−0.9158=0.2411

Finally, substituting the computed quantities above into Equation (7.23), we have the following:

cuv=c0.40.3=0.04840.2190∗0.2411=0.9160

To further illustrate the shape of the bivariate symmetric Kotz type density function, Figure 7.1 graphs the bivariate density function for the following:

1. 
   

   1. N = 2.0, s = 1.0, r = 0.5, ρ = 0.1N=2.0,s=1.0,r=0.5,ρ=0.1.

   
   
2. 
   

   2. N = s = 1, r = 0.5, ρ = 0.5N=s=1,r=0.5,ρ=0.5: bivariate normal distribution.

Figure 7.1 Copula density plots for Kotz type bivariate distribution.

Symmetric Bivariate Pearson Type VII Distribution

The PDF of symmetric bivariate Pearson type VII distribution can be given as follows:

fx1x2=N−1πm1−ρ21+1m1−ρ2×12+x22−2ρx1x2−N(7.24)

where N>1N>1, and m>0m>0 are parameters.

When N=m2+1, Equation (7.24) is the bivariate t-distribution with m degrees of freedom. When m=1,n=32, Equation (7.24) is the bivariate Cauchy distribution.

The marginal PDF of symmetric bivariate Pearson type VII distribution is as follows:

qx=Γn−12πmΓN−11+x2m−N−12(7.25)

The corresponding CDF of symmetric bivariate Pearson type VII distribution can be written as follows:

Qx=Γn−12πmΓN−1∫−∞x1+t2m−N−12dt(7.26)

where x ∈ (−∞, ∞), m>0, N>1.x∈−∞∞,m>0,N>1.

Then, the copula density function c(u, v)cuv can be given as follows:

cuv=ΓN−1ΓNΓN−1221+Q1−1um1+Q1−1vmN−121−ρ2[1+1m1−ρ2Q1−1u2+Q1−1v2−2ρQ1−1uQ1−1vN(7.27)

Example 7.5 Show the following bivariate Pearson type VII distribution cases are true.

Show the following cases are true:

1. 
   

   1. N=m2+1, the bivariate Pearson type VII distribution is the bivariate Student t-distribution with m degrees of freedom.

   
   
2. 
   

   2. m=1,N=32, the bivariate Pearson type VII distribution is the bivariate Cauchy distribution.

Solution:

1. 
   

   1. N=m2+1

   
   

   When N=m2+1, the probability density function generator of the Pearson type VII distribution may be rewritten as follows:

   
   
   g2PVII=Γm2+1Γm2−1πm1+tm−m2+1;N>1,m>0,N=m2+1
   
   
   
   

   Comparing with the probability density function generator for the bivariate Student t-distribution g2t=πv−1Γv2+1Γv21+tv−v2+1, we show that when N=m2+1, the bivariate Pearson type VII distribution reduces to the bivariate Student t-distribution. The same conclusion is reached for higher-dimensional cases.

   
   
2. 
   

   2. m=1,N=32

   
   

   When m=1,N=32, the probability density function generator of the Pearson type VII distribution may be rewritten as follows:

   
   
   g2PVIIt=Γ32Γ12π1+t−32,m=1,N=32
   
   
   
   

   Comparing with the probability density function generation for the Cauchy distribution g2Cauchy=π−1Γ32Γ121+t−32, we show that when m=1,N=32, the bivariate Pearson type VII distribution reduces to the bivariate Cauchy distribution. The same conclusion is reached for higher-dimensional cases.

Example 7.6 Compute the Pearson type VII copula density with the information given as follows: m = 0.5,  N = 2.0,  ρ = 0.1, u = 0.4, v = 0.3m=0.5,N=2.0,ρ=0.1,u=0.4,v=0.3.

Solution: Applying Equation (7.26), we can compute Q⁻¹(u), Q⁻¹(v)Q−1u,Q−1v numerically as follows:

Q⁻¹(u) =  − 0.1443; Q⁻¹(v) =  − 0.3086

Q−1u=−0.1443;Q−1v=−0.3086

Substituting

Q⁻¹(u) =  − 0.1443; Q⁻¹(v) =  − 0.3086Q−1u=−0.1443;Q−1v=−0.3086

into Equation (7.24), we can compute the copula density function c(0.4, 0.3)c0.40.3 as c(0.4, 0.3) = 1.1941c0.40.3=1.1941.

To illustrate the shape of Pearson type VII distribution, we graph the Pearson type VII copula density function for the following parameters in Figure 7.2:

1. 
   

   1. m = 0.5,  N = 2.0,  ρ = 0.1m=0.5,N=2.0,ρ=0.1.

   
   
2. 
   

   2. m = 3,  N = 2.5,  ρ = 0.2m=3,N=2.5,ρ=0.2: bivariate Student t-distribution with degrees of freedom as 3.

   
   
3. 
   

   3. m=1,N=32,ρ=0.2: bivariate Cauchy distribution.

Figure 7.2 Pearson type VII copula density plots.

Symmetric Bivariate Pearson Type II Distribution

The PDF of symmetric bivariate Pearson type II distribution can be expressed as

fx1x2=m+1π1−ρ21−x12+x22−2ρx1x21−ρ2m,∀x1x2Σ−1x1x2T≤10otherewise(7.28)

where m> − 1m>−1.

The marginal PDF can be given as follows:

qx=Γm+2πΓm+321−x2m+12;x∈−11(7.29)

The corresponding CDF can be given as follows:

Qx=Γm+2πΓm+32∫−1×1−t2m+12dt;x∈−11(7.30)

The copula probability density function can then be given as follows:

cuv=m+1Γ2m+32Γ2m+21−11−ρ2Q−1u2+Q−1v2−2ρQ−1uQ−1vm1−ρ21−Q−1u2m+121−Q−1v2m+12(7.31)

Example 7.7 Compute the bivariate Pearson type II copula density function with information given as follows: m =  − 0.5, ρ = 0.1, u = 0.4, v = 0.3m=−0.5,ρ=0.1,u=0.4,v=0.3.

Solution: Applying Equation (7.30), we can compute Q⁻¹(0.4), Q⁻¹(0.3)Q−10.4,Q−10.3 numerically as follows:

Q⁻¹(0.4) =  − 0.2; Q⁻¹(0.3) =  − 0.4

Q−10.4=−0.2;Q−10.3=−0.4

Applying Equation (7.31), we can compute the bivariate Pearson type II copula density function as follows:

c(0.4, 0.3) = 0.7091

c0.40.3=0.7091.

To further illustrate the shape of the bivariate Pearson type II copula density function, we graph the Pearson type II copula density function for the following parameters in Figure 7.3:

• 
  

  m =  − 0.5, ρ = 0.1m=−0.5,ρ=0.1; (2) m = 0.5, ρ = 0.2m=0.5,ρ=0.2.

Figure 7.3 Pearson type II copula density function plots.

7.2.1 Meta-Gaussian Copula

A d-dimensional meta-Gaussian copula can be expressed as follows:

Cu1…udΣ=ΦΣΦ−1u1…Φ−1ud=∫−∞Φ−1u1…∫−∞Φ−1ud12πd2Σ12exp−12wTΣ−1wdw(7.32)

where Φ⁻¹(⋅)Φ−1⋅ represents the inverse function of standard normal distribution; Φ_Σ(Φ⁻¹(u₁), …., Φ⁻¹(u_d))ΦΣΦ−1u1….Φ−1ud represents multivariate standard normal distribution function; ΣΣ represents the correlation matrix; Σ=1⋯ρ1d⋮⋱⋮ρd1⋯1,

ρij=1i=jρjii≠j,ρij=sinπτij2,τi,j the rank correlation coefficient;

d the dimension of continuous multivariate random variables; and w the integral matrix: w = [w₁, …, w_d]^Tw=w1…wdT.

Let gw1…wd=12πd2Σ12exp−12wTΣ−1w,x1=Φ−1u1,…,xd=Φ−1ud, Equation (7.32) may be rewritten as follows:

Cu1…ud=∫−∞x1⋯∫−∞xdgw1…wddw1…dwd(7.32a)

and its copula density function can be given as follows:

cu1…udΣ=∂d∂u1…∂udCu1…udΣ=∂d∂u1…∂ud∫−∞Φ−1u1⋯∫−∞Φ−1ud12πd2Σ12exp−12wTΣ−1wdw(7.33)

or equivalently

cu1…udΣ=∂d∂u1…∂ud∫−∞x1⋯∫−∞xdgw1…wddw1…dwd(7.33a)

Applying the partial derivative rule of inverse function,

u1=Φx1…ud=Φxd⇒du1dx1=ϕx1…duddxd=ϕxd⇒∂x1∂u1=dx1du1=1du1dx1=1ϕx1…∂xd∂ud=dxddud=1duddxd=1ϕxd(7.34)

In Equation (7.34), Φ(⋅)Φ⋅ is the CDF of standard normal distribution: Φx=∫−∞x12πe−t22dt; and ϕ(⋅)ϕ⋅ is the PDF of standard normal distribution: ϕx=12πe−x22.

Now substituting Equation (7.34) back into Equation (7.32) or (7.32a), we can calculate the partial derivatives for the d-dimensional meta-Gaussian copula in what follows.

First-Order Partial Derivative

Using ∂C∂u1 as an example, the first-order partial derivative of the meta-Gaussian copula may be derived as follows:

∂C∂u1=∂∂u1∫−∞x1…∫−∞xdgw1…wddw1…dwd=∂∂x1∫−∞x1…∫−∞xdgw1…wddw1…dwd∂x1∂u1=1ϕx1∫−∞x2…∫−∞xdgx1…wddw2…dwd(7.35)

Second-Order Partial Derivative

Using ∂2C∂u1∂u2 as an example, the second-order partial derivative of the meta-Gaussian copula may be derived as follows:

∂2C∂u1∂u2=∂∂u2∂C∂u1=∂∂u21ϕx1∫−∞x2…∫−∞xdgx1…wddw2…dwd=∂∂x21ϕx1∫−∞x2…∫−∞xdgx1…wddw2…dwd∂x2∂u2=1ϕx1ϕx2∫−∞x3…∫−∞xdgx1x2…wddw3…dwd(7.36)

dth-Order Partial Derivative

Repeating the derivative d-times, we obtain the meta-Gaussian copula density function as follows:

cu1…udΣ=∂du1…udΣ∂u1…∂ud=1ϕx1…ϕxdgx1…xd(7.37)

Let ς = [x₁, …, x_d]^T = [Φ⁻¹(u₁), …, Φ⁻¹(u_d)]^Tς=x1…xdT=Φ−1u1…Φ−1udT. Equation (7.37) may be rewritten as follows:

cu1…udΣ=112πe−x122…12πe−xd22·12πd2Σ12exp−12ςTΣ−1ς

=112πe−Φ−1u122…12πe−Φ−1ud22·12πd2Σ12exp−12ςTΣ−1ς

=1∏i=1d12πe−Φ−1ui22·12πd2Σ12exp−12ςTΣ−1ς

=Σ−12·1∏i=1de−Φ−1ui22exp−12ςTΣ−1ς(7.38)

Note that in Equation (7.38), ∏i=1de−Φ−1ui22 may be rewritten as follows:

∏i=1de−Φ−1ui22=exp−Φ−1u12+…+Φ−1ud2(7.38a)

Φ−1u12+…+Φ−1ud2=Φ−1u1…Φ−1udΦ−1u1…Φ−1ud=ςTς(7.38b)

Substituting Equations (7.38a) and (7.38b) into Equation (7.38), Equation (7.38) may be simplified as follows:

cu1…udΣ=Σ−121e−ςTς2exp−12ςTΣ−1ς=Σ−12exp−12ςTΣ−1ς+ςTς2(7.39)

Recall that ς^Tς = ς^TIςςTς=ςTIς, where I is d by d identity matrix. Equation (7.39) may also be rewritten as follows:

cu1…udΣ=Σ−12exp−12ςTΣ−1−Iς(7.39a)

Example 7.8 (Bivariate meta-Gaussian copula): Compute the bivariate meta-Gaussian copula and its copula density function with the given information: Σ=10.20.21,u1=0.4,u2=0.3, and show the first-order derivative of the bivariate meta-Gaussian copula.

Solution: Applying Equation (7.32) for d = 2d=2, we have the bivariate meta-Gaussian copula as follows:

Cu1u2Σ=ΦΣΦ−1u1Φ−1u2=∫−∞Φ−1u1∫−∞Φ−1u212πΣ12exp−12x1x2Σ−1x1x2Tdx1dx2(7.40)

From standard normal distribution, we have the following:

Φ⁻¹(0.4) =  − 0.2533; Φ⁻¹(0.3) =  − 0.5244

Φ−10.4=−0.2533;Φ−10.3=−0.5244.

Σ−1=1.0417−0.2083−0.20831.0417;Σ=0.96.

Substituting Φ⁻¹(0.4), Φ⁻¹(0.3), Σ⁻¹Φ−10.4,Φ−10.3,Σ−1 and ∣Σ∣∣Σ∣ into Equation (7.40), we have the following:

C(0.4, 0.3; Σ) = 0.1474

C0.40.3Σ=0.1474.

Applying Equation (7.39a) for d = 2d=2, we have the meta-Gaussian copula density function as follows:

cu1u2Σ=Σ−12exp−12Φ−1u1Φ−1u2Σ−1−IΦ−1u1Φ−1u2;I=1001=Σ−12exp−12x1x2Σ−1−Ix1x2(7.40a)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. 3 – Copulas and Their Properties
2. 10 – Rainfall Frequency Analysis
3. 13 – Drought Analysis
4. 12 – Water Quality Analysis