Using epishiny with linelist data
Sebastian Funk, Paul Campbell
Source:vignettes/articles/linelist_data.Rmd
linelist_data.RmdThe epishiny package can work with pre-aggregated data or with line lists, i.e. data sets in the form of tables that contain cases as one line per individual.
The package provides a range of visualisations of different aggregations of the data that can either be launched as individual modules or as part of a shiny dashboard that can be run locally or deployed on a server.
In this demonstation we will walk throught the steps of preparing external data then visualise it by launching individual modules from within an R script.
Load in data
The package comes with a built in example line list
df_ll, but a user can also bring their own data. Here we
will use a line list of Ebola in Sierra Leone published in Fang et al. (2016).
suppressPackageStartupMessages(library("readr"))
suppressPackageStartupMessages(library("dplyr"))
suppressPackageStartupMessages(library("purrr"))
suppressPackageStartupMessages(library("sf"))
suppressPackageStartupMessages(library("epishiny"))
url <- paste(
"https://raw.githubusercontent.com/parksw3/epidist-paper/main/data-raw/",
"pnas.1518587113.sd02.csv",
sep = "/"
)
df <- read_csv(url)
#> Rows: 8358 Columns: 8
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ","
#> chr (6): Name, Sex, Date of symptom onset, Date of sample tested, District, ...
#> dbl (2): ID, Age
#>
#> 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)
#> Rows: 8,358
#> Columns: 8
#> $ ID <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,…
#> $ Name <chr> "*****", "*****", "*****", "*****", "*****", "…
#> $ Age <dbl> 20, 42, 45, 15, 19, 55, 50, 8, 54, 57, 50, 27,…
#> $ Sex <chr> "F", "F", "F", "F", "F", "F", "F", "F", "F", "…
#> $ `Date of symptom onset` <chr> "18-May-14", "20-May-14", "20-May-14", "21-May…
#> $ `Date of sample tested` <chr> "23-May-14", "25-May-14", "25-May-14", "26-May…
#> $ District <chr> "Kailahun", "Kailahun", "Kailahun", "Kailahun"…
#> $ Chiefdom <chr> "Kissi Teng", "Kissi Teng", "Kissi Tonge", "Ki…Set up geo data
We next need geological data if we want to show maps.
geoBoundaries host sub-national administrative boundary shapefiles in various formats for every country in the world on github.
The line list contains Districts (admin 2 level) and Chiefdoms (admin 3 level). We can download the corresponding data sets from geoBoundaries via the following URLs.
shape_paths <- list(
adm2 = "https://github.com/wmgeolab/geoBoundaries/raw/905b0ba/releaseData/gbHumanitarian/SLE/ADM2/geoBoundaries-SLE-ADM2_simplified.geojson",
adm3 = "https://github.com/wmgeolab/geoBoundaries/raw/905b0ba/releaseData/gbHumanitarian/SLE/ADM3/geoBoundaries-SLE-ADM3_simplified.geojson"
)
shapes <- map(shape_paths, st_read)
#> Reading layer `geoBoundaries-SLE-ADM2_simplified' from data source
#> `https://github.com/wmgeolab/geoBoundaries/raw/905b0ba/releaseData/gbHumanitarian/SLE/ADM2/geoBoundaries-SLE-ADM2_simplified.geojson'
#> using driver `GeoJSON'
#> Simple feature collection with 14 features and 5 fields
#> Geometry type: MULTIPOLYGON
#> Dimension: XY
#> Bounding box: xmin: -13.30893 ymin: 6.923379 xmax: -10.27056 ymax: 9.999253
#> Geodetic CRS: WGS 84
#> Reading layer `geoBoundaries-SLE-ADM3_simplified' from data source
#> `https://github.com/wmgeolab/geoBoundaries/raw/905b0ba/releaseData/gbHumanitarian/SLE/ADM3/geoBoundaries-SLE-ADM3_simplified.geojson'
#> using driver `GeoJSON'
#> Simple feature collection with 167 features and 5 fields
#> Geometry type: MULTIPOLYGON
#> Dimension: XY
#> Bounding box: xmin: -13.30901 ymin: 6.923379 xmax: -10.27056 ymax: 9.999253
#> Geodetic CRS: WGS 84
# little bit of cleaning on district names to match with data
shapes$adm2$shapeName <- gsub("Area ", "", shapes$adm2$shapeName)
map(shapes, head)
#> $adm2
#> Simple feature collection with 6 features and 5 fields
#> Geometry type: MULTIPOLYGON
#> Dimension: XY
#> Bounding box: xmin: -13.30893 ymin: 7.340922 xmax: -10.27056 ymax: 9.999253
#> Geodetic CRS: WGS 84
#> shapeName shapeISO shapeID shapeGroup shapeType
#> 1 Kailahun 92492822B20773800307824 SLE ADM2
#> 2 Kenema 92492822B72931932323564 SLE ADM2
#> 3 Kono 92492822B77765797687755 SLE ADM2
#> 4 Bombali 92492822B76753077936535 SLE ADM2
#> 5 Kambia 92492822B57757993837772 SLE ADM2
#> 6 Koinadugu 92492822B53659457212824 SLE ADM2
#> geometry
#> 1 MULTIPOLYGON (((-10.80489 7...
#> 2 MULTIPOLYGON (((-10.85505 8...
#> 3 MULTIPOLYGON (((-10.64344 8...
#> 4 MULTIPOLYGON (((-12.32046 8...
#> 5 MULTIPOLYGON (((-13.2635 8....
#> 6 MULTIPOLYGON (((-10.58741 9...
#>
#> $adm3
#> Simple feature collection with 6 features and 5 fields
#> Geometry type: MULTIPOLYGON
#> Dimension: XY
#> Bounding box: xmin: -11.02525 ymin: 7.758386 xmax: -10.27056 ymax: 8.505881
#> Geodetic CRS: WGS 84
#> shapeName shapeISO shapeID shapeGroup shapeType
#> 1 Dea 93885176B88884577288656 SLE ADM3
#> 2 Jawie 93885176B40623889028536 SLE ADM3
#> 3 Kissi Kama 93885176B21354692554384 SLE ADM3
#> 4 Kissi Teng 93885176B65469407152324 SLE ADM3
#> 5 Kissi Tongi 93885176B56250152587449 SLE ADM3
#> 6 Kpeje Bongre 93885176B52534318264914 SLE ADM3
#> geometry
#> 1 MULTIPOLYGON (((-10.60592 7...
#> 2 MULTIPOLYGON (((-10.91103 7...
#> 3 MULTIPOLYGON (((-10.43962 8...
#> 4 MULTIPOLYGON (((-10.43962 8...
#> 5 MULTIPOLYGON (((-10.39938 8...
#> 6 MULTIPOLYGON (((-10.8119 8....Launch Modules
Before we launch the modules we can define some grouping variables.
If passed to the time or place modules, a
select input will appear in the ‘options’ dropdown allowing you to group
the data by the variables you select.
If you want a more readable variable label to appear in the module rather than the variable name itself, pass a named vector where the name is the label and the value is the variable name.
In our example we use sex and district as a variable, but since they are already formatted as labels there is no need to pass names:
group_vars <- c("Sex", "District")Place module
Now that we have the shapefiles we can collate the information
contained in the format that epishiny expects using the
geo_layer() function. Since we are using more than one geo
layer, we combine them in a list:
geo_data <- list(
geo_layer(
layer_name = "District",
sf = shapes$adm2,
name_var = "shapeName",
join_by = c("shapeName" = "District")
),
geo_layer(
layer_name = "Chiefdom",
sf = shapes$adm3,
name_var = "shapeName",
join_by = c("shapeName" = "Chiefdom")
)
)We use this to launch the place module:
launch_module(
module = "place",
df = df,
geo_data = geo_data,
group_vars = group_vars[1] # only pass sex variable since district is already visualised on map
)If you select the Chiefdom admin 3 level from the options dropdown menu you will see a warning meassage informing you that more that 50% of the cases could not be matched to the shapefile, so some matching of disparate place names would be required in this case. We won’t do that here but if you need help with this task check out our hmatch package.
Time module
To launch the time module, we need to pass the date variable(s) in the line list we want to use for the x-axis.
Notice that the date variables are of character class in the data but they are automatically parsed to date class in the time module via the lubridate::as_date function.
launch_module(
module = "time",
df = df,
date_vars = c("Date of symptom onset", "Date of sample tested"),
group_vars = group_vars
)Person module
Finally, we can plot an age/sex pyramid using the person module, passing age and sex variable names and the levels in the sex variable to encode male and female:
launch_module(
module = "person",
df = df,
age_var = "Age",
sex_var = "Sex",
male_level = "M",
female_level = "F"
)