Reputation: 21
I need to open 100 ndjson large files (with same columns) , I have prepared a script to apply to each file but I would not like to repeat this 100 times !
With ndjson::stream_in , I can only open 1 ndjson file into R as a data frame
I tried the process to open multiple csv files and consolidate them into 1 dafatframe only, but it does not work with ndjson files :(
library(data.table)
library(purrr)
map_df_fread <- function(path, pattern = "*.ndjson") {
list.files(path, pattern, full.names = TRUE) %>%
map_df(~fread(., stringsAsFactors = FALSE))
}
myfiles <-
list.files(path = "C:/Users/sandrine/Documents/Projet/CAD/A/",
pattern = "*.ndjson",
full.names = T) %>%
map_df_fread(~fread(., stringsAsFactors = FALSE))
I tried to find also a package to convert ndjson files into csv ...but did not find any.
Any idea?
Upvotes: 2
Views: 208
Reputation: 4698
Using your own approach that you mentioned first, does this work?
library(tidyverse)
library(ndjson)
final_df <-
list.files(path = "C:/Users/sandrine/Documents/Projet/CAD/A/",
pattern = "*.ndjson",
full.names = T) %>%
map_dfr(~stream_in(.))
Upvotes: 1