Description Usage Arguments Details References See Also Examples
Fit the parameters of a peak over threshold (POT) model.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 | fitPot(x, ...)
## S3 method for class 'data.frame'
fitPot(obj, ...)
## S3 method for class 'matrix'
fitPot(obj, ...)
## S3 method for class 'formula'
fitPot(form, x, ...)
## S3 method for class 'numeric'
fitPot(x, dt = NULL, u = 0, method = "mle",
declust = NULL, r = 1, rlow = 0.75, nboot = 1000,
declust.scale = FALSE)
## S3 method for class 'fpot'
coef(obj, rate = FALSE, ci = FALSE, alpha = 0.05)
## S3 method for class 'fpot'
vcov(obj, rate = FALSE)
## S3 method for class 'fpot'
print(obj)
## S3 method for class 'fpot'
predict(obj, rt = c(2, 5, 10, 20, 50, 100), se = FALSE,
ci = FALSE, alpha = 0.05, nboot = 0, ...)
|
x |
Sample. |
dt |
Date or time of observation. |
u |
Threshold. |
method |
Estimation method. Either |
declust |
Method for declustering. If necessary, Either 'run' or 'flood'. |
r |
Lag parameter for declustering. Either the run parameter
or the minimum time between two flood Peaks.
The scale must coincide with the observation date |
rlow |
Declustering parameter. Percentage that define a minimal recession time between two flood peaks. |
nboot |
Number of bootstrap sample. Used to extimate the covariance
matrix with the method of L-moments and predict return period.
For prediction, if |
se, ci |
Should the Standard error or the confident interval be return. The standard error are obtained by the delta method and the confident interval by nonparametric boostrap. |
The access functions coef
and vcov
return respectively the
parameters and the variance-covariance matrix of the POT model. The
variance covariance matrix is available only with the method 'mle'
The access function predict
returns the return periods.
If dt
is a Date the return period is computed in year using the range
of observation.
The rate
(i.e. number of event per year) can be manually adjusted
in case. By default rate = 1
.
The declustering can be a simple 'run' declustering, i.e. a cluster
start when passing a threshold and stop when r
steps are below
the threshold.
Choulakian V, Stephens MA. (2001) Goodness-of-Fit Tests for the Generalized Pareto Distribution. Technometrics. 43(4):478–84.
Coles S. (2001) An introduction to statistical modeling of extreme values. Springer Verlag.
Davison AC, Smith RL. (1990) Models for Exceedances over High Thresholds. Journal of the Royal Statistical Society Series B (Methodological). 52(3):393–442.
which.floodPeaks, mrlPlot.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 | xd <- rgpa(100, 1, -.2)
fit <- fitPot(xd, u = 0)
print(fit)
vcov(fit)
predict(fit)
fit <- fitPot(flow~date, canadaFlood$daily, u = 1000,
declust = 'flood', r = 14)
print(fit)
plot(flow~date,canadaFlood$daily, type = 'l', col = 4)
points(fit$time,fit$excess+fit$u, col = 2)
predict(fit, se = TRUE,ci = TRUE)
fit <- fitPot(flow~date, canadaFlood$daily, u = 1000,
declust = 'flood', r = 14, method = 'lmom')
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.