As I'm starting my career in analytics, I have to choose between SAS in-memory analytics and Openware and largely adopted R, which one should I choose now so that I will have good market value in short-term as well as long-term?

  • $\begingroup$ What do you mean "have to choose"? Are you talking about your learning focus, what platform you're going to be adopting for your company, what software you're going to use for self-study or classes, or what? Makes a big difference. $\endgroup$
    – Wayne
    Feb 28, 2016 at 18:16
  • $\begingroup$ Can anyone answer this question? - datascience.stackexchange.com/questions/12619/… $\endgroup$
    – Minu
    Jul 11, 2016 at 16:57

5 Answers 5


You'll need both or even more for your "career in analytics". Start with SAS, SAS studio is intuitive and easy, once you are comfortable with the "stats" and looking for next level than is coming R in place.

SAS also have free courses for intro in stat and their platform. SAS free courses

For long term believe me that is not about the tools/platform. You can check the free book listed bellow and most important modeling and prediction techniques listed in the book. Theoretical and practical understanding of many important methods is essential and after that you will be able to compute in any language.

An Introduction to Statistical Learning


Go for R. SAS is a monster and is not fun. Having said that your market value will be determined more by what you can and want to do and the general education level than just by this or that technology.


This all depends. SAS is great and is backed by the SAS Institute, meaning if you're working for an organization that has invested in SAS, you can contact the support team for anything funky happening with the software.

R is free and open source, and there are also organizations being created that are building on to and supporting R much like SAS Institute has done. The difference being that SAS has been around much longer and is much more structured than R.

They both have their ups and downs, and it depends on what type of analytics you plan on doing in your career. So, to answer your questions, I would decide what you envision yourself doing. To my understanding R is good for Big Data applications and Machine Learning, SAS is great for statistical analysis (ARIMA, etc).

SAS and R comparisons:




In my experience, there are a few things to consider here:

  • What field you're going into
  • What kinds of technology you believe you'll be working with
  • What kinds of teams you believe you'll be working with


This is a huge determiner. To my understanding, SAS is standard in finance, banking, some biostats, and other industries. R, on the other hand, is open source, free, and is receiving a lot of attention currently from Microsoft after their acquisition of Revolution Analytics.


Are you thinking you'll be at a large corporation or a small shop? Do you think you'll need to work with creating production ready algorithms or simply producing insights and analysis?

The reason this matters is that open source technology can often times be a bit easier to convert to production ready use. The cost of software can also be a barrier at smaller companies and may dictate what you are capable of using.

Collaboration with Other Teams

If you sit firmly on the business side, it's possible that everyone you work with may use SAS or may be comfortable with the outputs. In this case, you may not need to collaborate across any additional technology.

If you work with technology, however, it may be difficult to integrate SAS or proprietary solutions into other workflows. An example would be creating a real-time user scoring system. If you were able to program this into R or Python, you may be able to pass the code directly to a developer for implementation. This would be more difficult if a proprietary solution was involved.

Other Considerations

The analytics space is evolving rapidly. On top of the above, Python is coming out as a very popular technology for use in machine learning and data mining and pairs well with Spark as well as some other large data technology. A specific example of a library here would be scikit-learn.

Closing Thoughts/My Experience

I'm an R/Python user by experience, though I have some experience with SAS in school as well as in work. Generally, I've found the level of support with R and Python to be great - especially since they've started to rise in popularity. SAS has its uses and has definitely carved out piece of the industry for itself.

All in all, however - getting a solid base in statistical theory, programming (scripting), and understanding the application and value that analysis can provide will go a long way. The syntax between these tools is, generally, not worlds apart. SAS can be a bit strange compared to R and Python, but all of these tools generally have syntax which is fairly readable and is not difficult to adapt to another tool.


@ImperativelyAblative has many good points.

Adding to that, many companies, especially banks, that rooted in SAS have started to experiment with R. This is because of the cost saving and maturing enterprise support, think of RStudio Server and Cloudera. SAS is great, but it comes with a premium price tag (some may argue it's even overpriced).

I am a management consultant working at a top tier firm with specialty in data science. We serve most of the global companies and work with their executives and operating teams. Moving to open source tools, such as R, has gained the blessing of many top executives and building tractions among core analytics teams (Marketing, Modelling, etc.). The point being - the market is moving towards open source tools

The other consideration, in my opinion, is the role you aspire to play in an organization. Here are some archetypes based on my experience in working with the analytics teams:

  • General Business Analyst: R, Interactive visualization tools (Tableau / QlikView), some SQL, and Excel / Powerpoint (of course)
  • Modeller: SAS is must, R (my clients are all trying to pick this up), deep in SQL / HQL (querying on Hadoop stack), and strong domain knowledge (risk model, pricing, operation optimization, etc.)
  • Application Developer (people who put things into production): SAS is must, Python, and automation language

Of course, this is not an exhaustive list, but hope it provides some industry trend for your reference. There is always the superstar in any organization who knows all these stuff, could be you :)


Not the answer you're looking for? Browse other questions tagged or ask your own question.