lambda_L = lim_{u->0+} C(u,u) / u. The lower tail of (X, Y) is the
upper tail of (-X, -Y), so the estimator is Coles' chi(u) applied to
the reflected sample; no second estimator is written. For a radially
symmetric dependence structure the two coefficients coincide, which is
the anchor used in the tests.
Usage
Tldepl(y, copula, theta = 0.95)
Arguments
- y
Numeric vector, first margin.
- copula
Numeric vector, second margin (equal length).
- theta
Threshold in (0, 1), default 0.95.
Value
List with estimate, u, n, method.
References
Joe, H. (1997). Multivariate Models and Multivariate
Dependence Concepts. Chapman and Hall, section 2.1.10.
Examples
Tldepl(1:20, 1:20, 0.5)
#> $estimate
#> [1] 1
#>
#> $u
#> [1] 0.5
#>
#> $n
#> [1] 20
#>
#> $method
#> [1] "lower tail dependence via reflected empirical chi(u) [Joe 1997]"
#>
