The default structure of a {cards} object is data frame. But it is often needed to represent the results in other formats.
The {cards} objects can easily be expressed in YAML and JSON formats.
Let’s begin by creating an Analysis Results Data (ARD) object.
library(cards)
ard <-
bind_ard(
ard_continuous(ADSL, by = "ARM", variables = "AGE"),
ard_categorical(ADSL, by = "ARM", variables = "AGEGR1")
)
ard
#> {cards} data frame: 51 x 11
#> group1 group1_level variable variable_level stat_name stat_label stat
#> 1 ARM Placebo AGE N N 86
#> 2 ARM Placebo AGE mean Mean 75.209
#> 3 ARM Placebo AGE sd SD 8.59
#> 4 ARM Placebo AGE median Median 76
#> 5 ARM Placebo AGE p25 Q1 69
#> 6 ARM Placebo AGE p75 Q3 82
#> 7 ARM Placebo AGE min Min 52
#> 8 ARM Placebo AGE max Max 89
#> 9 ARM Xanomeli… AGE N N 84
#> 10 ARM Xanomeli… AGE mean Mean 74.381
#> ℹ 41 more rows
#> ℹ Use `print(n = ...)` to see more rows
#> ℹ 4 more variables: context, fmt_fn, warning, errorYAML
The as_nested_list() function prepares the ARD data
frame to be converted to a YAML object.
ard |>
as_nested_list() |>
yaml::as.yaml() |>
cat()
#> variable:
#> AGE:
#> group1:
#> ARM:
#> group1_level:
#> Placebo:
#> stat_name:
#> 'N':
#> stat: 86
#> stat_fmt: '86'
#> warning: ~
#> error: ~
#> context: continuous
#> mean:
#> stat: 75.2093023
...JSON
The as_nested_list() function prepares the ARD data
frame to be converted to a JSON object.
ard |>
as_nested_list() |>
jsonlite::toJSON(pretty = TRUE)
#> {
#> "variable": {
#> "AGE": {
#> "group1": {
#> "ARM": {
#> "group1_level": {
#> "Placebo": {
#> "stat_name": {
#> "N": {
#> "stat": [86],
#> "stat_fmt": ["86"],
#> "warning": {},
#> "error": {},
#> "context": ["continuous"]
#> },
...Use the jsonlite::write_json() function to convert to
JSON and write to disk simultaneously.
REST API
With the JSON format, ARDs can be made accessible via REST APIs.
Using the {plumber} package, an API can be created in a
plumber.R file.
library(cards)
library(plumber)
library(jsonlite)
#* @get /nested_json
#* @serializer unboxedJSON
function() {
ard <- bind_ard(
ard_continuous(ADSL, by = "ARM", variables = "AGE"),
ard_categorical(ADSL, by = "ARM", variables = "AGEGR1")
)
ard |>
as_nested_list() |>
toJSON()
}The following code serves the API, making it available for testing.
Note: “plumber.R” represents the path to the plumber.R file
created above.
Data is requested from a hosted API via the {httr} package and converted back to a nested list using {jsonlite}.