Comparison Study of SAS Vs R

06-11-2013
Source By: 

Will R out perform SAS on Data Analytics in near future?A comparison study of SAS vs. R:

 Data Analytics means SAS and SAS means data analytics. Can any analytics firm think analytics services without SAS in recent years? But for last couple of years people have started talking about a data analytics tool call “R”.

What is R?

R is anopen source data analytics tool which was created by Ross Ihaka and Robert Gentleman  in 1993.

It is a complete interactive and object-oriented programming language with Data Visualization, Statistical Analysis,Predictive Modeling, Data Mining, and Machine Learning algorithms

Revolution Analytics focused on developing “open-core” versions of the free and open source software R for enterprise, academic and analytics customers

Case Study: SAS and R benchmarking (Source:Big Data Analytics, Benchmarking SAS®, R, and Mahout)

The objective of the comparison study is required to understand the quality of the analytics, overall completeness, and overall effort is required by the analytics to do the analysis and the cost effectiveness without compromising the quality. Let us take a use case where we want to build a predictive model by three methods 1) Logistics Regression, 2) Decision Tree and 3) Ransom Forest. SAS has done a comparison study on the same and here is the result:

Findings:

Based on the study

  • SAS is providing the efficient and powerful solution of on the three prediction modeling methods spending additional money to buy the required licenses.
  • Still SAS would be more accurate and effortless tools to use for analytics and specific solutions in Retail, Fraud, BI, Text mining and others.
  • R could be a cost effective optionbut required skills to write the code, build the solution and validate the results which is itself another time consuming efforts

Recent Market Trend:

  • The TIOBE Community Programming Index ranks the popularity of programming languages ranked R in 24th place and SAS at 31st, in January 2012.
  • The Transparent Language Popularity Index ranks R in 14th place and SAS, in 27th in May, 2013
  • Very few are considering a switch to R from SAS

Please note that the result could be biased as the sample size is very small and the segment could be from R community.

Conclusion:

In summary, based on the latest industry trends R and open-source technologies are becoming the standard for statistical computing over the next ten years. This trend has created multiple new opportunities to develop the product or solution development using open source platforms in analytics like R.

Based on the Gartner’s research, industry still believes that SAS is the leader and the only tool which can deliver end to end analytics with high quality, provide different domain specific solutions and industry does not mind to spend extra money to maintain the business standard at any cost.



Comments: 1


  • Deepali

    Hi Sandip,

    I had this question for a long time in my mind..Having been in BI and associated with Analytics and Data mining I saw very few customers go big bang for Data Mining…Most of them were still doing their implementation of DWH, Data Integrity, MDM , Reporting, Governance. Very few visionary companies really took to Data Mining on a serious Note., Couldn’t agree more that SAS Predictive/Statistical Models are surely time tested….But I guess with customer bending towards the cloud infrastrusture they have learnt to store the (in-house valuables) like Data, outside their infrastructure,,,,Probably hence the broader acceptance to open source technologies like R. Jasper, Brio have been around for near decade but customers are willing to experiment…not sure is this because they don’t want to get bogged with product upgrades and probably want their space to experiment without much of license hassles…But yes strange when people talk Analytics we hear more of R than SAS….Is it just a bend of mindset towards some new buzz word or old wine new bottle ???

FOLLOW US ON

LATEST POSTS

  • Subscribe for Blog Updates

  • TAG CLOUD

  • ARCHIVES