Thx all !good question actuaryalfred !genkideska, yeah thats what I needed. Is there a genrator phi which generates Frechet bounds ?

<t>Is there a Independent copula in the family of Archimedean copulas? I know that the product copula: Pi(u,v)=u*v is an independent, I was asked if there is any Archimedean independent copula. I dont know, cant find the answer. And what about the Frechet upper and lower bounds, do they correspond t...

Anyone out there who interested in dependence concepts (correlations, rank correlations, copulas) who would drop few references on that subject? I can only find Nelsen's Introduction to copulas, is there anything else, more general about dependence modelling. thanks

Thanks a lot for your usefull thoughts. Even though it seems that representing joint distribution in terms of copulas is the best way of measure the dependence.

hello, we all know how great copulas are to measure the dependence (because they allow us to choose marginal distributions freely and so on), but what are the disadvantages of this approach ???? it has to have some pitfalls! so what are the weak points of copulas ??Thanks a lot M

%%double post

yeahhh it works !!!! THANKS THANKS THANKS !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!omg thats fantastic! THANKSSSSSSSSSSSSSSSSSSSSSSSSSSS

<t>hmmmmmmm thank you very much. ... but what ??? sorry Im being very dump.I have found the following code to generate the plot of normal copula distribution which I understand. n=500;Z = mvnrnd([0 0], [1 .8; .8 1], n);U = normcdf(Z,0,1); plot(U(:,1),U(:,2),'.');So what you saying is to use Z=mvnpdf...

anyone ??? someone ??? do u know how to denerate the density of normal copula ???? very important to methx

@Jasba this is taken from High Dimensional Dependence Modelling !! x,y,z are in fact rank correlations. and the function f(x,y,z) is the numerator of the expression for partial correlation.

@ppauper thanks a lot for your answer and the time u spent on it! much appraciated I still would have to think about that sinus transformatin I dont think I got it

Originally posted by: lballabioQuoteYou need a four-dimensional screen.ooh bammer, I have only a flat screen!!! lol Quote You have to attack the problem mathematically.Mathematics is my weak point!!!

<t>QuoteOriginally posted by: ppauperhey, the chlorine should bleach your hair ! hahaha it does ! I can never dye my hair cos the chlorine wahses the color out. QuoteJust looking at it, I can only see "z" in one place so solve for "z" for the case f=0f(x,y,z)=0=x-y.*z.*sqrt((1-x.^2).*(1-y.^2))+x.*y....

I think you have to use copulas in here. Copula 'joins' together marginals.the joint density f_{1234} looks like:f_{1234}=f_1*f_2*f_3*f_4*c(F_1,F_2,F_3,F_4)where f_i are densities of X_i and F_i are the distributions and c a copula!

<t>Hi ! I am not blond but I my brain doesn't differ much from blond girl's brain, it is only smaller. I think I spend too much time swimming all this chlor is eating all my gray cells, there is no space left where I can provide mathematical computations. Anyway I happen to do master thesis in Maths...