Getting and cleaning data - JSON

Question 1

JSON SALIENT FACTS

Answer 1

Javascript Object Notation

Lightweight data storage

Common format for data from application programming interfaces (APIs)

Similar structure to XML but different syntax/format

Data stored as

Numbers (double)

Strings (double quoted)

Boolean (true or false)

Array (ordered, comma separated enclosed in square brackets [])

Object (unorderd, comma separated collection of key:value pairs in curley brackets {})

Question 2

JSON PACKAGE & FIRST STEPS

Answer 2

install.packages(“jsonlite”)

library(jsonlite)

reading data into memory

jsonData <- fromJSON(“https://api.github.com/users/jtleek/repos”)

Question 3

JSON EXPLORING DATA

Answer 3

Use loading variable as handle to refer to data in memory

Find 1st level headings:
> names(jsonData)

Drill to secon level by referring to a chosen heading using $
> names(jsonData$owner)

List content
> jsonData$owner$login

Question 4

WRITING DATA FRAME TO JSON

Answer 4

Start with a DF assigned to the variable iris

Transform DF into a JSON format
> myjson <- toJSON(iris, pretty=TRUE)

Read the result
> cat(myjson)

Convert back from JSON
> iris2 <- fromJSON(myjson)

Display result
> head(iris2)

Question 5

JSON RESOURCES

Answer 5

http://www.json.org/

A good tutorial on jsonlite :
http://www.r-bloggers.com/new-package-jsonlite-a-smarter-json-encoderdecoder/