In pvisser82/earthquakedata: Visualise Earthquake Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 7

)
library(earthquakedata)
library(dplyr)
library(ggplot2)
library(leaflet)

Aim of the package

The earthquakedata package allows the user to clean and visualise the earthquake data set provided by the U.S. National Oceanographic and Atmospheric Administration (NOAA). This data set contains information about 5,933 earthquakes over an approximately 4,000 year time span.

Download here

The earthquakedata package has 6 functions that are exported for use by users and one internal function that is not visible:

eq_clean_data()
geom_timeline()
geom_timeline_label()
theme_timeline()
eq_map()
eq_create_label()

Clean the data

To be able to visualise the data set properly, it is necessary to clean the data first.

This is achieved by:

Creating a DATE column by uniting YEAR, MONTH and DAY and converting it to the Date class
Converting the LATITUDE and LONGITUDE columns to numeric class
Stripping out the country name (including colon) from the LOCATION_NAME column and converting it to title case

clean_data <- readr::read_delim("../earthquakes.tsv.gz",delim = "\t") %>%
  eq_clean_data()

Display the earthquake timeline

The easiest way to display the earthquake data for multiple countries over a large time period is by using a timeline.

To assist the user with this, the geom_timeline() function can be used in conjunction with the ggplot2 package. The geom_timeline() function uses the following aesthetics - please note that the user is able to change this to other columns if required

x = DATE
y = COUNTRY
size = EQ_PRIMARY
colour = DEATHS

readr::read_delim("../earthquakes.tsv.gz",delim = "\t") %>%
  eq_clean_data() %>%
  dplyr::filter(COUNTRY == c("USA") & lubridate::year(DATE) >= 2000) %>%
  ggplot() +
    geom_timeline(aes(x = DATE, y = COUNTRY,size = EQ_PRIMARY, colour = DEATHS))

The above sample code will display a very basic graph using the default ggplot2 theme. To allow the user to see some more information and to make the graph a bit more pleasing to the eye, the geom_timeline_label() and theme_timeline() functions were added to the package.

The geom_timeline_label() function adds a vertical line with a text label to identify specific earthquakes easily. To manage the number of labels that will be plotted, the n_max parameter is used. It will limit the number of labels to the n_max amount and will show the labels for the earthquake events with the highest magnitude.

The theme_timeline() function is a modified version of theme_classic() and does the following:

The y axis line, ticks and title are hidden
The y axis line text is changed to darkgray with size 11 and a right-hand margin of 15
The x axis line colour black with a thickness of 1
The x axis text is changed to darkgray with size 11 with a top margin of 3
The x axis title is size 14 with a top margin of 5
The plot a title is added and justified horisontally in the middle of the plot
The legend is positioned at the bottom of the screen

Below is an example that reads the data set, cleans it, filters it to specific countries and dates, plots it, labels the top 5 events and uses the correct theme. Please note that the labels for the legend are added as well along with a title for the plot:

readr::read_delim("../earthquakes.tsv.gz",delim = "\t") %>%
  eq_clean_data() %>%
  dplyr::filter(COUNTRY == c("USA","MEXICO") & lubridate::year(DATE) >= 2000) %>%
  ggplot() +
    geom_timeline(aes(x = DATE, y = COUNTRY,size = EQ_PRIMARY, colour = DEATHS)) +
    geom_timeline_label(aes(x = DATE, y = COUNTRY, label = LOCATION_NAME, size = EQ_PRIMARY), n_max = 5) +
    ggtitle("Earthquake Timeline") +
    theme_timeline() +
    labs(size = "Richter Scale value:", colour = "# of Deaths:")

Timeline

Map the earthquake data

Using the leaflet package, it is possible to plot the earthquake information on an interactive map. The user can use the eq_map() function to plot the earthquakes on a map. Each earthquake is indicated by a blue circle. The size of the circle on the map is relative to the magnitude of the earthquake it represents. The annot_col parameter used by the eq_map function reflects a column in the data set that must be displayed when the user clicks on a specific earthquake on the map. The default value is DATE

 readr::read_delim("earthquakes.tsv.gz", delim = "\t") %>%
  eq_clean_data() %>%
  dplyr::filter(COUNTRY == "MEXICO" & lubridate::year(DATE) >= 2000) %>%
  eq_map(annot_col = "DATE")

Basic plot

The eq_create_label() function is used to provide even more information. It takes the data set as an argument and creates an HTML label that can be displayed on the map as a pop-up. The label consists of the Location, Magnitude and Total deaths. If one of these fields are missing a value, the field is ignored when the label is built

readr::read_delim("earthquakes.tsv.gz",delim = "\t") %>%
  eq_clean_data() %>%
  dplyr::filter(COUNTRY == "MEXICO" & lubridate::year(DATE) >= 2000) %>%
  dplyr::mutate(popup_text = eq_create_label(.))%>%
  eq_map(annot_col = "popup_text")