Provides a flexible visualization of the country-level Covid-19 spread,
inpired by the displays created by John Burn-Murdoch from the Financial
Times. Uses data from the Johns Hopkins University CSSE team
(https://github.com/CSSEGISandData/COVID-19), the ACAPS governmental
measures database
(https://www.acaps.org/covid19-government-measures-dataset), and the
World Bank (https://data.worldbank.org).
If your are overwhelmed with the options of the plot, explore them with
shiny_covid19_spread() and use the 'copy code for plot to clipboard'
option.
plot_covid19_spread(
data = download_merged_data(cached = TRUE, silent = TRUE),
type = "deaths",
min_cases = ifelse(per_capita, ifelse(type == "deaths", 5, 50), ifelse(type ==
"deaths", 1000, 10000)),
min_by_ctry_obs = 7,
edate_cutoff = 180,
data_date_str = format(lubridate::as_date(data$timestamp[1]), "%B %d, %Y"),
cumulative = TRUE,
change_ave = 7,
per_capita = FALSE,
per_capita_x_axis = ifelse(per_capita, TRUE, FALSE),
population_cutoff = 0,
log_scale = TRUE,
highlight = NULL,
exclude_others = FALSE,
intervention = NULL
)The data frame to base the plot on. Should be a merged data
frame obtained by download_merged_data and defaults to
download_merged_data(cached = TRUE, silent = TRUE).
The statistic that you want to plot. Needs to be either "confirmed", "deaths", "recovered" or "active", defined as the difference of "confirmed", "deaths" and "recovered".
Defines the zero point of your X axis (the 'event date'). Defaults to 1,000 cases for deaths and 10,000 cases otherwise.
Limits the plot to countries that have at least that many days of data since and including the event date. Defaults to 7.
The upper limit of the X axis in event days. Defaults to 180.
A date string to include in the annotation of the plot giving the time when the data was pulled. Defaults to the timestamp of the data. Note that you might run into issues with the default when running this in a non-english locale. Consider setting it by hand then.
If TRUE (the default) data is being plotted as
cumulative (showing the total figures). If FALSE, (averaged) daily
changes are plotted instead. See change_ave to set the averaging
window.
Number of days to average over when you plot daily changes.
If TRUE data is being plotted as per capita measure
based on World Bank data. Defaults to FALSE.
If TRUE, the 'event date' cutoff for the x
axis set by min_cases is evaluated based on by capita
measures (cases per 100,000 inhabitants). Only feasible when
per_capita is TRUE. Other than in older versions
it now defaults to TRUE whenever per_capita is TRUE
else FALSE.
Do you want to restrict the plot to countries that exceed a certain population cutoff? Takes a value in millions and defaults to 0. Useful for per capita displays.
Do you want the Y-axis to be log-scaled? Defaults to
TRUE.
A character vector of ISO3c (ISO 3166-1 alpha-3) codes that
identify countries that you want to highlight. Using the
gghighlight package, these observations are highlighted by color
and labeled while the other are grayed out. In NULL (the default),
all countries are labeled. This can cause very cluttered plots.
If TRUE, unhighlighted countries are excluded
from the plot. If FALSE (the default), unhighlighted countries
are grayed out.
If not default NULL then this identifies the
intervention type that you want to be highlighted by a point marker.
Valid intervention types are based on the ACAPS government measure data
and include general lockdowns ('lockdown'), social distancing
('soc_dist'), movement restrictions ('mov_rest'), public health measures
('pub_health'), and social and economic measures ('soc_econ').
A ggplot2 object.
# NOT RUN {
plot_covid19_spread()
merged <- download_merged_data(cached = TRUE, silent = TRUE)
plot_covid19_spread(merged, highlight = "DEU", intervention = "lockdown")
plot_covid19_spread(merged, type = "recovered", min_by_ctry_obs = 10,
edate_cutoff = 40,
highlight = c("ITA", "ESP", "FRA", "DEU", "USA"), intervention = "soc_dist"
)
# }
Run the code above in your browser using DataLab