Plots of sample L-Skewness ans L-Kurtosis estimates at various
thresholds for peaks over threshold modelling, using the Generalized
Pareto parametrization.
Usage
lmomplot(data, u.range, nt = max(50, length(data)), identify = TRUE,
...)
Arguments
data
A numeric vector.
u.range
A numeric vector of length two, giving the limits for
the thresholds at which the model is fitted.
nt
The number of thresholds at which the sample L-moments are
evaluated.
identify
Logical. If TRUE, points on the plot are
identify using identify function.
...
Other arguments to be passed to the model fit
function fitgpd.
Warnings
L-moments plot are really difficult to interpret. It can help us to
say if the GP distribution is suited to model data.
Author
Mathieu Ribatet
Details
For each thresholds, sample L-skewness and L-kurtosis are computed. If
data are GP distributed, one have :
$$ \tau_4 = \frac{\tau_3 \left( 1 + 5 \tau_3 \right)}{5 + \tau_3}
$$
So, a threshold is acceptable if sample \(\left(\tau_3,
\tau_4\right)\) are near the theoretical curve.
References
Hosking, J. R. M. and Wallis, J. R. (1997) Regional
Frequency Analysis. Cambridge University Press.
Begueria, S. (2005) Uncertainties in partial duration series
modelling of extremes related to the choice of the threshold value.
Journal of Hydrology, 303(1-4): 215--230.