| Library: | VaR |
| See also: | VaRcharfDG |
| Quantlet: | VaRcgfDG | |
| Description: | computes the cumulant generating function (cgf) for the class of quadratic forms of Gaussian vectors. |
| Usage: | s = VaRcgfDG(t,par) | |
| Input: | ||
| t | the complex argument to the cgf | |
| par | a list defining the distribution; contains at least the following components: theta - the constant term delta - the linear term lambda - the diagonal of the quadratic term | |
| Output: | ||
| s | the value of the cgf at t | |
library("plot")
library("VaR")
library("math")
proc() = VaRcgfDGtest(par,n,xlim)
dt =(xlim[2]-xlim[1])/(n-1)
t = xlim[1] +(0:(n-1))*dt
r = VaRcgfDG(complex(t,t*0),par)
z1 = setmask(t~r.re, "line", "blue")
plot(z1)
endp
theta = 0
delta = #(1)
lambda= #(0.1)
par = list(theta,delta,lambda)
VaRcgfDGtest(par,100,#(-2,2))
Plots the cumulant generating function for a distribution, which is close to normal distribution. The graph is close to a parabola.