Using epishiny with aggregated data
Source:vignettes/articles/aggregated_data.Rmd
aggregated_data.RmdEpidemiological surveillance data is often reported in an aggregated form by ministries of health. A typical aggregation would be by a geographical and time unit, cases and deaths by health area and week for example.
You can use epishiny to visualise data in this form by
declaring one or more count_vars in the data (numeric
columns containing the aggregation count totals).
Let’s run through an example using WHO’s COVID-19 daily cases and
deaths dataset. The data contains daily case and death totals per
country, so we can visualise both the time and place component using
epishiny.
Import aggregated COVID-19 data from WHO
df_who_covid <- read_csv("https://covid19.who.int/WHO-COVID-19-global-data.csv")
#> Rows: 50160 Columns: 8
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ","
#> chr (3): Country_code, Country, WHO_region
#> dbl (4): New_cases, Cumulative_cases, New_deaths, Cumulative_deaths
#> date (1): Date_reported
#>
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
glimpse(df_who_covid)
#> Rows: 50,160
#> Columns: 8
#> $ Date_reported <date> 2020-01-05, 2020-01-12, 2020-01-19, 2020-01-26, 202…
#> $ Country_code <chr> "AF", "AF", "AF", "AF", "AF", "AF", "AF", "AF", "AF"…
#> $ Country <chr> "Afghanistan", "Afghanistan", "Afghanistan", "Afghan…
#> $ WHO_region <chr> "EMRO", "EMRO", "EMRO", "EMRO", "EMRO", "EMRO", "EMR…
#> $ New_cases <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 6, 17, 67, 183, 247, 3…
#> $ Cumulative_cases <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 7, 24, 91, 274, 521, 9…
#> $ New_deaths <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 3, 10, 15, 13…
#> $ Cumulative_deaths <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 5, 15, 30, 43…Import country boundaries with rnaturalearth
world_map <- ne_countries(scale = "small", type = "countries", returnclass = "sf") %>%
st_transform(crs = 4326) %>%
select(iso_a2_eh, name, pop_est)
# setup the geo layer for epishiny
geo_data <- geo_layer(
layer_name = "Country",
sf = world_map,
name_var = "name",
pop_var = "pop_est",
join_by = c("iso_a2_eh" = "Country_code")
)
geo_data
#> $layer_name
#> [1] "Country"
#>
#> $sf
#> Simple feature collection with 177 features and 5 fields
#> Geometry type: MULTIPOLYGON
#> Dimension: XY
#> Bounding box: xmin: -180 ymin: -90 xmax: 180 ymax: 83.64513
#> Geodetic CRS: WGS 84
#> First 10 features:
#> iso_a2_eh name pop_est geometry
#> 1 FJ Fiji 889953 MULTIPOLYGON (((180 -16.067...
#> 2 TZ Tanzania 58005463 MULTIPOLYGON (((33.90371 -0...
#> 3 EH W. Sahara 603253 MULTIPOLYGON (((-8.66559 27...
#> 4 CA Canada 37589262 MULTIPOLYGON (((-122.84 49,...
#> 5 US United States of America 328239523 MULTIPOLYGON (((-122.84 49,...
#> 6 KZ Kazakhstan 18513930 MULTIPOLYGON (((87.35997 49...
#> 7 UZ Uzbekistan 33580650 MULTIPOLYGON (((55.96819 41...
#> 8 PG Papua New Guinea 8776109 MULTIPOLYGON (((141.0002 -2...
#> 9 ID Indonesia 270625568 MULTIPOLYGON (((141.0002 -2...
#> 10 AR Argentina 44938712 MULTIPOLYGON (((-68.63401 -...
#> lon lat
#> 1 177.97595 -17.9376200
#> 2 34.14207 -6.2078294
#> 3 -12.57202 24.2305626
#> 4 -110.24381 56.7019200
#> 5 -99.31483 37.2367450
#> 6 66.31159 48.0689612
#> 7 63.44288 41.3532772
#> 8 144.22612 -6.6678356
#> 9 113.26946 -0.1785159
#> 10 -64.08055 -37.2391995
#>
#> $name_var
#> [1] "name"
#>
#> $pop_var
#> [1] "pop_est"
#>
#> $join_by
#> iso_a2_eh
#> "Country_code"
#>
#> attr(,"class")
#> [1] "epishiny_geo_layer"Define count variables in the data
We are only insterested in the new case and death variables, since the time module will handle calculating cumulative numbers for us. Here we supply a named vector to show different variable labels in the module’s indicator select input.
count_vars <- c("Cases" = "New_cases", "Deaths" = "New_deaths")Launch time module
launch_module(
module = "time",
df = df_who_covid,
date_vars = "Date_reported",
group_vars = "WHO_region",
count_vars = count_vars,
show_ratio = TRUE,
ratio_lab = "CFR",
ratio_numer = "New_deaths",
ratio_denom = "New_cases",
date_intervals = c("week", "month", "year")
)Launch person module
The COVID-19 data has no age or sex variables so we can’t use the person module, but for demonstation purposes we will show this can also be used with an aggregated data set below.
# create a data set with case and death counts aggregated by age group and sex
age_levels <- c("<5", "5-17", "18-24", "25-34", "35-49", "50+")
sex_levels <- c("Male", "Female")
df_as <- tibble(
sex = factor(c(rep(sex_levels[1], 6), rep(sex_levels[2], 6))),
age_group = factor(rep(age_levels, 2), levels = age_levels),
cases = round(runif(12, 20, 100)),
deaths = round(runif(12, 0, 20)),
)
df_as
#> # A tibble: 12 × 4
#> sex age_group cases deaths
#> <fct> <fct> <dbl> <dbl>
#> 1 Male <5 63 17
#> 2 Male 5-17 29 13
#> 3 Male 18-24 96 3
#> 4 Male 25-34 81 2
#> 5 Male 35-49 29 3
#> 6 Male 50+ 51 3
#> 7 Female <5 42 18
#> 8 Female 5-17 78 7
#> 9 Female 18-24 88 11
#> 10 Female 25-34 50 16
#> 11 Female 35-49 24 14
#> 12 Female 50+ 81 10
# launch the module passing age, sex and count_var info
launch_module(
module = "person",
df = df_as,
age_group_var = "age_group",
sex_var = "sex",
male_level = "Male",
female_level = "Female",
count_vars = c("cases", "deaths")
)