[TCopula] Spearman rho and moment matching.

[lrnv]

@Santymax98 If you want to take a shot at it, there is there an algorithm that matches both teh kendall tau and spearman rho to the df and rho of a bivariate student. Also they rpovide quick formulas for evaluating the spearman rho of the student, which we clearly need here...

One last thing to do would be to implement once and for all the multivariate student cdf, to finallyr educe computational time on this model. For that, see the mvtnorm R package which has links to the right literature (genz 2004 i think)

[Santymax98]

I'll check it out. I'm currently finishing the documentation. I also think we should specify the Rho for empirical copulas like Checker and Empirical because it takes too long to calculate.

Regarding the CDF, I think these are always done with Monte Carlo methods.

[lrnv]

I'll check it out. I'm currently finishing the documentation.

Take your time :)

I also think we should specify the Rho for empirical copulas like Checker and Empirical because it takes too long to calculate.

Yes, and a few others too. I was trying to setup benchmarks to highlight a list of potential methods to improve, but it take too long to compute i wa not yet able to get an exploitable result :P

Regarding the CDF, I think these are always done with Monte Carlo methods.

yes, but a proper monte-carlo adapted to the model, like the one implemented in the link i gave, will be much faster than what we have now.

[Santymax98]

Okay, let's try it.

[lrnv]

See also this one for fitting : jds-online.org/journal/JDS/article/1266/info