Reputation: 32986
In trying to analyze tweets from a hashtag, I am able to obtain 100 results at a time via this call.
curl -i "http://otter.topsy.com/search.json?q=esa2012&window=a&perpage=100&offset=0" >> esa2012_0.json
resulting in a file like this (JSON file in link).
How do I read this into R
and convert it to a data.frame
?
Thus far I am able to skip the first several (non-JSON) lines like so:
library(XML)
library(RJSONIO)
file0 <- scan(file = "~/Desktop/data/esa2012_0.json", skip = 18, what= "raw")
but the resulting read becomes difficult to coerce.
Upvotes: 1
Views: 413
Reputation: 72769
fromJSON(file0)
should do it.
I typically run it with ,simplify=FALSE
and construct the data.frame myself rather than trying to have it simplify for me.
Note you don't need the command line call to curl
if you use the RCurl library:
library(RJSONIO)
library(RCurl)
j <- getURL("http://blah?mine=yours")
fromJSON(j)
Upvotes: 3