---
title: "Calculating and Visualising Exceedances"
author: "Robert W Schlegel"
date: "`r Sys.Date()`"
description: "This vignette demonstrates the use of the exceedence function."
output:
rmarkdown::html_vignette:
fig_caption: yes
vignette: >
%\VignetteIndexEntry{Calculating and Visualising Exceedances}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
bibliography: bibliography.bib
---
```{r global_options, include = FALSE}
knitr::opts_chunk$set(fig.width = 8, fig.height = 3, fig.align = 'centre',
echo = TRUE, warning = FALSE, message = FALSE,
eval = TRUE, tidy = FALSE)
```
## Exceedances
In addition to the calculation of events based on a given percentile, this package also allows for the calculation of events based on a given static threshold with the `exceedance()` function. This is most useful if one is more interested in testing a time series for events that exceed a pre-determined static threshold that is already known to be of particular significance in a given part of the world. The output of this function may also be used in the calculation of events that may require two thresholds, such as atmospheric events with 'tMin' and 'tMax' daily values. For a detailed explanation for how to do this please see [this vignette](https://robwschlegel.github.io/heatwaveR/articles/complex_clims.html). The data requirements for `exceedance()` are the same as for `ts2clm()` and `detect_event()`.
## Calculating exceedances
The calculation of exceedances may be done with only one function as seen here:
```{r exceedance-example1}
# Activate libraries
library(dplyr)
library(ggplot2)
library(heatwaveR)
# Calculate exceedence
exc_25 <- exceedance(sst_WA, threshold = 25)
# Look at a few metrics
exc_25$exceedance %>%
ungroup() %>%
select(exceedance_no, duration, date_start, date_peak, intensity_max, intensity_cumulative) %>%
dplyr::arrange(-intensity_cumulative) %>%
head(5)
```
Note that the resultant output of `exceedance()` is very similar to `detect_event()`, except that the two dataframes within the list are called `threshold` and `exceedance`, rather than `climatology` and `event`.
## Visualising exceedances
Because `event_line()` and `lolli_plot()` are designed to work on the output of `detect_event()`, if we want to visualise the results of `exceedance()` we will need to do so 'by hand' with __`ggplot2`__. The code below works as an example of how to do so:
```{r fig-example1, echo = TRUE, eval = TRUE}
exc_25_thresh <- exc_25$threshold %>%
slice(9500:9866)
ggplot(data = exc_25_thresh, aes(x = t)) +
geom_flame(aes(y = temp, y2 = thresh, fill = "all"), show.legend = F) +
geom_line(aes(y = temp, colour = "temp")) +
geom_line(aes(y = thresh, colour = "thresh"), size = 1.0) +
scale_colour_manual(name = "Line Colour",
values = c("temp" = "black", "thresh" = "forestgreen")) +
scale_fill_manual(name = "Event Colour", values = c("all" = "salmon")) +
guides(colour = guide_legend(override.aes = list(fill = NA))) +
scale_x_date(date_labels = "%b %Y") +
labs(y = expression(paste("Temperature [", degree, "C]")), x = NULL)
```
<!--  -->
## Calculating negative exceedances
Unfortunately there is no antonym for exceedance in the English language, which makes talking about exceedances _below_ a static threshold somewhat awkward. For the purposes of clarity here we will refer to these as negative exceedances.
The `exceedance()` function may be used to calculate consecutive days below a threshold as seen here:
```{r exceedance-example2}
exc_19 <- exceedance(sst_WA, threshold = 19, below = TRUE)
exc_19$exceedance %>%
dplyr::ungroup() %>%
dplyr::select(exceedance_no, duration, date_start, date_peak, intensity_mean, intensity_cumulative) %>%
dplyr::arrange(intensity_cumulative) %>%
head(5)
```
## Visualising negative exceedances
Were one to desire a visualisation of these data it could be produced with the following code:
```{r fig-example2, echo = TRUE, eval = TRUE}
exc_19_thresh <- exc_19$threshold %>%
slice(1500:1866)
ggplot(data = exc_19_thresh, aes(x = t)) +
geom_flame(aes(y = thresh, y2 = temp), fill = "steelblue3", show.legend = F) +
geom_line(aes(y = temp, colour = "temp")) +
geom_line(aes(y = thresh, colour = "thresh"), size = 1.0) +
scale_colour_manual(name = "Line Colour",
values = c("temp" = "black", "thresh" = "forestgreen")) +
scale_y_continuous(limits = c(18, 23.5)) +
scale_x_date(date_labels = "%b %Y") +
labs(y = expression(paste("Temperature [", degree, "C]")), x = NULL)
```
<!--  -->