Fetches graph data directly from Neo4j REST API into R. Useful for R igraph users.

neo4R_example.R

# Requirements
#sudo apt-get install libcurl4-gnutls-dev # for RCurl on linux
#install.packages('RCurl')
#install.packages('RJSONIO')

library('RCurl')
library('RJSONIO')
 
query <- function(querystring) {
  h = basicTextGatherer()
  curlPerform(url="http://localhost:7474/db/data/cypher",
    postfields=paste('query',curlEscape(querystring), sep='='),
    writefunction = h$update,
    verbose = FALSE
  )
  
  result <- fromJSON(h$value())
  
  data <- data.frame(t(sapply(result$data, unlist)))
  names(data) <- result$columns
  
  data 
}
 

# EXAMPLE
# =======

# Cypher Query:
q <- "match (o:Organization)-[r]-(p:Person) return o.name,o.location,p.account,p.name,p.email limit 20"

data <-query(q)

head(data,20)

# Output:
#           o.name                          o.location      p.account         p.name            p.email
# 1     PerfectLine                             Estonia         kritik                                  
# 2      Sappho OSS                           London UK    andrewheald                                  
# 3          The 88                                 NYC         aface1                                  
# 4          The 88                                 NYC     xbilldozer                                  
# 5          The 88                                 NYC         chadyj                                  
# 6          The 88                                 NYC       benmanns Benjamin Manns benmanns@gmail.net
# 7        simplabs                     Munich, Germany        marcoow                                  
# 8   Everyday Hero                 Brisbane, Australia    soloman1124                                  
# 9   Everyday Hero                 Brisbane, Australia         orodio                                  
# 10  Everyday Hero                 Brisbane, Australia justinhennessy                                  
# 11  Everyday Hero                 Brisbane, Australia           coop     Tim Cooper coop@latrobest.org
# 12  Everyday Hero                 Brisbane, Australia      evilmarty   Marty Zalega    marty@zalega.co
# 13 Sorenson Media  Salt Lake City, UT & San Diego, CA       bcarlson                                  
# 14 Sorenson Media  Salt Lake City, UT & San Diego, CA      elmomalmo                                  
# 15 Sorenson Media  Salt Lake City, UT & San Diego, CA        enthooz                                  
# 16         3scale Barcelona, Spain and Sunnyvale, USA          solso                                  
# 17         3scale Barcelona, Spain and Sunnyvale, USA   MarkCheshire                                  
# 18         3scale Barcelona, Spain and Sunnyvale, USA          rhoml                                  
# 19         3scale Barcelona, Spain and Sunnyvale, USA           mikz  Michal Cichra                   
# 20         3scale Barcelona, Spain and Sunnyvale, USA           njyx                                  

_Caution:_ This script runs into trouble if the query response is in a non-tabular format or includes missing values. Neo4j does not support null-valued properties and any property with a null value is removed from the database. In other words running SET node.property = NULL elicits the same result as running REMOVE node.property, which means it deletes the property. Neo4j's REST API responds to queries in JSON, which is schema-less as well, and does not represent missing properties in any way. When converting the JSON output to a tabular format, this results in rows with variable number of columns. In case your database contains missing values, either replace them with symbolic values or convert them on the fly within your query. If none of these is an option, you need to use a tabular format like CSV for data interchange.