Let's say I have this data:
df <- data.frame(value = c("foo", "bar", "baz"))
A simple toJSON gets me this:
[{"value":"foo"},{"value":"bar"},{"value":"baz"}]
But if I need the values as single element arrays, how do I structure the original data? I've tried a few variations involving c(), list(), and I(). Nothing works to get me this output:
[{"value":["foo"]},{"value":["bar"]},{"value":["baz"]}]
To get the arrangement you want, you need a list of lists
jsonlite::toJSON(list(list("value"="foo"), list("value"="bar"), list("value"="baz")))
# [{"value":["foo"]},{"value":["bar"]},{"value":["baz"]}]
We can do this with a little help from the tidyverse function mutate_all to convert all the columns to lists.
library(tidyverse)
df %>% mutate_all(as.list) %>% jsonlite::toJSON()
# [{"value":["foo"]},{"value":["bar"]},{"value":["baz"]}]
value column, use mutate_at instead of mutate_all: df %>% mutate_at(vars(value), as.list) %>% jsonlite::toJSON().I() and setting auto_unbox to true.
unbox. It's always easier to help with a more realistic example.You can pass list to toJSON function.
library(jsonlite)
df <- data.frame(value = c("foo", "bar", "baz"))
# Iterate over rows and put them to list
# Output passed to toJSON output
toJSON(apply(df, 1, as.list))
# [{"value":["foo"]},{"value":["bar"]},{"value":["baz"]}]
Like this toJSON input is:
apply(df, 1, as.list)
[[1]]
[[1]]$value
[1] "foo"
[[2]]
[[2]]$value
[1] "bar"
[[3]]
[[3]]$value
[1] "baz"
apply converts the data.frame to a matrix so if you have columns of different datatypes, this might get messy. But for the simple case in the question where there is just one column, this is very nice.toJSON to treat the data as it's presented.