heather_dplyr.R

library(tidyverse)
library(jsonlite)

x <- tribble(
  ~name, ~city, ~age,
  "Heather", "Renton", 31,
  "Jacqueline", "Issaquah", 34
)

# BAD WAY
toJSON(x, dataframe = "columns")


# GOOD WAY

# if you have like 100 columns to combine there is a way to do it without writing it out, but it takes longer
x %>%
  mutate(person = paste0(name, ", ", city,", ", age)) %>%
  select(person) %>%
  toJSON()

# JSONL WAY ( this is not the most efficient but it is the closest to)
result <-
  x %>%
  transpose() %>% # this turns it from a dataframe to a list where each element is a row, which makes this all easier
  map_chr(toJSON, auto_unbox=TRUE) %>% # use map to turn each object into json individually, the auto_unbox is like... a thing I can explain over the phone
  paste0(collapse="\n") # combine them into a single string with newlines between each

cat(result) # just print it prettier if you are copying and pasting
write_file(result,"temp.jsonl") # if you want to save it