Update/correct GOF tests for vines
ulf85 opened this issue · comments
The GOF tests for vine copulas implemented in the package return so far only a correct test statistics. Some also do return a p-value. But these p-values are highly experimental and its calculation not yet checked.
Asymptotic p-values should be treated with high causious.
Further, some tests have bootstarpped p-values, but I never had the time to check if the C-code is working corretly.
I started with some corrections for the GOF test based on White for the bivariate case.
The needed variance-covariance matrix which is needed for the test statistic had a very poor approximation so far. I implemeted the variance-covariance matrix for the one-parameteric copulas.
Next will be the variance matrix for the t-copula.
Found another bug in BiCopGofTest. For the White test (method="white") always the asymptotic p-value is calculated independent of the parameter B, which should manage if a asymptotic p-value or a bootstrapped p-value is to be derived.
The other two tests seem to be ok.
Fix for the last bug in BiCopGofTest for method="white". Now bootstrapped p-values are returned if B is greater than 0.