EXPERIMENTAL: The function provides two themes of sampling distribution characterization by simulation of the first three L-comoment ratios (L-correlation \(\tau_{2[\ldots]}\), L-coskew \(\tau_{3[\ldots]}\) and L-cokurtosis \(\tau_{4[\ldots]}\)) of a copula. Subsequently, the sampling distribution can be used for inference.
First, semi-optional Monte Carlo integration estimation of the L-comoments of the parent copula are computed. Second, simulations involving the sample size \(n\) presumed the size of the actual sample from which the estimates of the sample L-comoments given as arguments. These simulations result in a report of the L-moments (not L-comoments) of the sampling distribution and these then are used to compute p-values for the L-comoment matrices provided by the user as a function argument.
lcomCOPpv(n, lcom, cop=NULL, para=NULL, repcoe=5E3, type="gno",
mcn=1E4, mcrep=10, usemcmu=FALSE, digits=5, ...)An R
list is returned.
A string functioning as a label for the remaining tables;
Another R list holding tables of the L-moments of the L-comoments derived from Monte Carlo integration for samples of size \(N =\) mcn. The simulations are replicated mcrep times; and
Another R list holding tables of the L-moments of the L-comoments derived from the small sample simulations for samples of size \(n =\) n as well as the p-values estimated by a generalized normal distribution (see lmomco package documentation) of the L-moments using either the small sample means or the mean of the replicated Monte Carlo integrations as dictated by usemcmu. In all circumstances, however, the results for the small sample simluations are tabulated in ntable only the p-value will be reflective the setting of usemcmu.
The sample size \(n\). This argument is semi-optional because \(n = 0\) can be given to skip corresponding simulations and the ntable on return will only contain NA; this feature permits rapid extraction of the Ntable and thus the lcom contents are simply not used;
The sample L-comoments (see below);
A copula function;
Vector of parameters, if needed, to pass to the copula;
The replication coefficient \(\phi\) affecting the number of simulations of size n;
The distribution type used for modeling the distribution of the sampling values. The generalized normal (see distribution type "gno" in package lmomco) accommodates some skewness compared to the symmetry of the normal ("nor") just in case situations arise in which non-ignorable skewness in the sample distribution exists. The distribution abbreviations of package lmomco are recognized for the type argument, but in reality the "nor" and "gno" should be more than sufficient;
The sample size \(N\) passed to the bilmoms function for the Monte Carlo integration. If \(N = 0\) then the Monte Carlo integration is not used, otherwise the minimum sample size is internally reset to \(N=4\) so that first four L-moments are computable;
The number of replications of the Monte Carlo simulation by bilmoms;
A logical toggling whether the mean value computed from the replicated Monte Carlo integrations is used instead of the mean values for the small sample simulation for the p-value computations;
The number of digits to round numerical entries in the returned tables and can be NA for no rounding; and
Additional arguments to pass to the bilmoms function or to the copula.
W.H. Asquith
The notation \(r[\ldots]\) refers to two specific types of L-comoment definitions and a blend between the two. The notation \(r[12]\) means that the \(r\)th L-comoment for random variables \(\{X^{(1)}, X^{(2)}\}\) where \(X^{(2)}\) is the sorted variable and \(X^{(1)}\) is shuffled by the sorting index. Conversely, the notation \(r[21]\) means that the \(r\)th L-comoment for random variables \(\{X^{(1)}, X^{(2)}\}\) where \(X^{(1)}\) is the sorted variable and \(X^{(2)}\) is shuffled by the sorting index. The notation \(r[12:21]\) means that the average between the \(r[21]\) and \(r[21]\) is computed, which might prove useful in circumstances of known or expected symmetry of the L-comoments.
Continuing, \(\hat\tau_{2[12]}\) is the sample L-correlation, \(\hat\tau_{3[12]}\) is the sample L-coskew, and \(\hat\tau_{4[12]}\) is the sample L-cokurtosis all with respect to the sorting of the second variable. The computation of these L-comoment matricies can be made by functions such as function lcomoms2() in the lmomco package. The number of replications for the simulations involving the \(n\) sample size is computed by
$$m = \phi/\sqrt{n}\mbox{,}$$
where \(\phi\) is the repcoe replication factor or coefficient. If usemcmu is TRUE then mcn \(> 0\) else usemcmu is reset to FALSE.
Nelsen, R.B., 2006, An introduction to copulas: New York, Springer, 269 p.
lcomCOP, COP, kullCOP, vuongCOP