Description Usage Arguments Details Value See Also Examples

View source: R/ithresh_methods.R

`plot`

method for class "ithresh". Produces an extreme value
threshold diagnostic plot based on an analysis performed by
`ithresh`

. Can also be used to produce a plot of
the posterior sample generated by `ithresh`

for a particular
training threshold.

1 2 3 4 5 6 7 8 9 10 11 12 |

`x` |
an object of class |

`y` |
Not used. |

`...` |
Additional arguments passed on to |

`which_v` |
A numeric scalar or vector. If If |

`prob` |
A logical scalar. If |

`top_scale` |
A logical scalar indicating Whether or not to add a scale
to the top horizontal axis. If this is added it gives the threshold on
the scale not chosen by |

`add_legend` |
A logical scalar indicating whether or not to add a
legend to the plot. If |

`legend_pos` |
The position of the legend (if required) specified using
the argument |

`which_u` |
Either a character scalar or a numeric scalar.
If If Otherwise, |

Produces plots of the *threshold weights*, defined in
equation (14) of
Northrop et al. (2017),
against training threshold. A line is produced for each of the validation
thresholds chosen in `which_v`

. The result is a plot like those in
the top row of Figure 7 in
Northrop et al. (2017).

It is possible that a curve on the plot may be incomplete. This indicates
that, for a particular threshold level, a measure of predictive
performance is `-Inf`

. This occurs when an observation in the data
lies above the estimated upper end point of the predictive distribution
produced when this observation is removed.

If `which_u`

is supplied then the object with which
`plot.evpost`

was called is returned (invisibly).
Otherwise, a list is returned (again invisibly) with two components.
`x`

is a vector containing the coordinates plotted on the
(lower) horizontal axis.
`y`

is an `length(u_vec)`

by `n_v`

matrix of
*threshold weights* obtained by normalising the columns of the
matrix `pred_perf`

returned by `ithresh`

.
See equation (14) of
Northrop et al. (2017).

`ithresh`

for threshold selection in the i.i.d. case
based on leave-one-out cross-validation.

`summary.ithresh`

Summarizing measures of threshold
predictive performance.

`print.ithresh`

Prints the threshold weights.

`predict.ithresh`

for predictive inference for the
largest value observed in N years.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | ```
# [Smoother plots result from making n larger than the default n = 1000.]
# Threshold diagnostic plot
u_vec_gom <- quantile(gom, probs = seq(0, 0.9, by = 0.05))
gom_cv <- ithresh(data = gom, u_vec = u_vec_gom, n_v = 3)
plot(gom_cv, lwd = 2, add_legend = TRUE, legend_pos = "topleft")
mtext("significant wave height / m", side = 3, line = 2.5)
# Plot of Generalized Pareto posterior sample at the best threshold
# (based on the lowest validation threshold)
plot(gom_cv, which_u = "best")
# See which threshold was used
summary(gom_cv)
# Plot of Generalized Pareto posterior sample at the highest threshold
n_u <- length(u_vec_gom)
plot(gom_cv, which_u = n_u, points_par = list(pch = 20, col = "grey"))
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.