Big Data/BI Zone is brought to you in partnership with:

Gary Sieling is a software developer interested in dev-ops, database technologies, and machine learning. He has a computer science degree from the Rochester Institute of Technology. He has worked on many products in the legal and regulatory industries, having worked on and supported several data warehousing applications. Gary is a DZone MVB and is not an employee of DZone and has posted 62 posts at DZone. You can read more from them at their website. View Full User Profile

Building a JSON Webservice in R

10.26.2012
| 4794 views |
  • submit to reddit

R is a programming language for mathematics and statistics. There are several R libraries available to support web development, including rjson and RJSONIO (note case – R library names are case sensitive). RJSONIO is based on rjson, but with modifications to improve performance working with large JSON payloads.

The example below returns the data required to render a scatter chart with a best fit line like this:

Chart in R-Studio

This requires that you’ve already set up rApache. To install RJSONIO, check http://cran.r-project.org/src/contrib/ to find the latest version, and use that filename in the following. It appears that CRAN removes the old versions- unfortunately they don’t have an easy way to just pull the latest.

wget http://cran.r-project.org/src/contrib/RJSONIO_1.0-1.tar.gz
R CMD INSTALL RJSONIO_1.0-1.tar.gz

The following code will pull JSON data from the URL, parse it, and return a linear model. This includes information for the graph, error bars, best fit line, axis intercepts, etc. For a charting package like ExtJS, the data may need to be re-formatted to render successfully, which fortunately R is quite good at.

setContentType("application/json")
 
library(RJSONIO)
library(utils)
 
cat(GET$data)
 
json <- fromJSON(URLdecode(GET$data))
fit <- lm(x ~ y)
 
cat(toJSON(fit))

Strictly speaking, this should be a POST, but this makes the example easier to read.

Some examples of this use the basicJSONHandler, like below. I found that this does not return numerics for JSON arrays, and opted not to use it.

h = basicJSONHandler()
x = fromJSON("[1, 2, 3]", h)

If you’ve read this far, you may enjoy my review of the R Cookbook.

This URL accepts and returns JSON results:

http://localhost:8081/R/fit.R?data={%22x%22:[1.5,2,7,8,15],%22y%22:[1.5,2,7,8,15]}

Results:

{ "assign" : [ 0,
      1
    ],
  "call" : { "" : "lm",
      "formula" : [ "~",
          "x",
          "y"
        ]
    },
  "coefficients" : { "(Intercept)" : 0.95408999999999999,
      "y" : 0.78297000000000005
    },
  "df.residual" : 3,
  "effects" : { "" : 0.23488999999999999,
      "(Intercept)" : -13.864000000000001,
      "y" : 8.5699000000000005
    },
  "fitted.values" : { "1" : 2.1284999999999998,
      "2" : 2.52,
      "3" : 6.4348999999999998,
      "4" : 7.2179000000000002,
      "5" : 12.699
    },
  "model" : { "x" : [ 1,
          3,
          6,
          9,
          12
        ],
      "y" : [ 1.5,
          2,
          7,
          8,
          15
        ]
    },
  "qr" : { "pivot" : [ 1,
          2
        ],
      "qr" : [ { "(Intercept)" : -2.2361,
            "y" : -14.981999999999999
          },
          { "(Intercept)" : 0.44721,
            "y" : 10.945
          },
          { "(Intercept)" : 0.44721,
            "y" : -0.17422000000000001
          },
          { "(Intercept)" : 0.44721,
            "y" : -0.26557999999999998
          },
          { "(Intercept)" : 0.44721,
            "y" : -0.90512999999999999
          }
        ],
      "qraux" : [ 1.4472,
          1.2826
        ],
      "rank" : 2,
      "tol" : 9.9999999999999995e-08
    },
  "rank" : 2,
  "residuals" : { "1" : -1.1285000000000001,
      "2" : 0.47997000000000001,
      "3" : -0.43489,
      "4" : 1.7821,
      "5" : -0.69865999999999995
    },
  "terms" : [ "~",
      "x",
      "y"
    ],
  "xlevels" : {  }
}

 

Published at DZone with permission of Gary Sieling, author and DZone MVB. (source)

(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)