Improved memory usage and RJSONIO compatibility in jsonlite 0.9.15

[This article was first published on OpenCPU, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

opencpu logo

The jsonlite package implements a robust, high performance JSON parser and generator for R, optimized for statistical data and the web. Last week version 0.9.15 appeared on CRAN which improves memory usage and compatibility with other packages.

Migrating to jsonlite

The upcoming release of shiny will switch from RJSONIO to jsonlite. To make the transition painless for shiny users, Winston Chang has added some compatibility options to jsonlite that mimic the (legacy) behavior of RJSONIO. The following wrapper results in the same output as RJSONIO::toJSON for the majority of cases. Hopefully this will make it easier for other package authors to make the transition to jsonlite as well.

# RJSONIO compatibility wrapper
toJSON_legacy <- function(x, ...) {
  jsonlite::toJSON(I(x), dataframe = "columns", null = "null", na = "null",
   auto_unbox = TRUE, use_signif = TRUE, force = TRUE,
   rownames = FALSE, keep_vec_names = TRUE, ...)
}

However be aware that the RJSONIO defaults can sometimes result in unexpected behavior and odd edge cases (which is why jsonlite was created in the first place). Therefore it is still recommended to switch to the jsonlite defaults when possible (see jsonlite paper for a discussion on the mapping). One exception is perhaps the auto_unbox argument, which many people seem to prefer to TRUE for encoding relatively simple static data structures.

Memory usage

The new version should use less memory when parsing JSON, especially from a file or URL. This is mostly due to a new push-parser implementation that can incrementally parse JSON in little pieces, which eliminates overhead of copying gigantic JSON strings. In addition, jsonlite now uses the new curl package for retrieving data via a connection interface.

mydata1 <- jsonlite::fromJSON("https://jeroenooms.github.io/data/dmd.json")

The call above is results in the same output as the call below, but it should consume less memory, especially for very large json files.

library(httr)
req <- GET("https://jeroenooms.github.io/data/dmd.json")
mydata2 <- jsonlite::fromJSON(content(req, "text"))

None of this changes anything in the API, these changes are all internal.

To leave a comment for the author, please follow the link and comment on their blog: OpenCPU.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)