There has been excitement building around the upcoming Jedox 5.1.With so many new features it could be called Jedox 6. One very powerful addition is the upcoming marriage of Jedox and R.
I’ve seen the enthusiasm among Jedox partners and clients, but what benefits does R bring to the business place of a Jedox user? As a prelude to Chris Mentor’s technical articles on Jedox 5.1 (starting soon), I speak here with Vladislav Malicevic, Head of R&D at Jedox, on what happens when the world’ s most innovative and easy to use BI solution meets probably the world’s most powerful statistical language.
Let’s start by looking at R.
Sam Perrin: What is R?
Vladislav Malicevic: R is a language and environment for statistical computing used for complex predictive analysis. With a community of over 2 million, you find R used from credit risk analysis in financial institutions, to reducing customer churn and optimizing your marketing spend.
Sam Perrin: Why R?
Vladislav Malicevic: Because R gives you the power to understand your business at a more fundamental level. R is true “analytics”, think of it as providing the “intelligence” in Business Intelligence. R is a huge step from simply getting your data into a cube to “slice and dice” or to visualizing your data in charts to manually search [SP: in the hope of noticing a trend with your eyes, see http://www.forbes.com/fdc/welcome_mjx.shtml].
R is very powerful at combining data into a cohesive model. If you know your customers’ credit card transactions, you could understand (and predict) how their taxi spend is affected by the weather. Imagine how this insight helps if you are disrupting an entire industry by allowing consumers to bid on fares [SP e.g. https://www.uber.com/], or for car-sharing services – or driverless cars.
Domain experts from around the world contribute to R and this make it extremely robust and applicable across many industries and business functions. Learning a statistical language means you need to put your thinking-cap on (we like to call this professional development), but if you have some fundamental stats knowledge, through forums and online help from the active user base means your learning curve is much shorter. You can even do university-level R courses for free. [SP: e.g. https://www.coursera.org/course/rprog https://www.udacity.com/course/ud651]
Another drawcard is with R’s price tag (3 minutes * and 50 megabytes), you can use this same functionality used by global financial institutions and some of the biggest internet companies) – out of the box with the standard premium Jedox suite.
Sam Perrin: Ok, tell me about the integration.
Vladislav Malicevic: R power users can now run any R scripts in Jedox and change data in R by writing back live through Jedox. They can customize charts and edit results through Jedox. Full R-functionality is available through all Jedox front-ends – mobile, MS Office, and the web. Jedox has always been easy to use, and now we are bringing that same intuitiveness and user-friendliness to R.
Sam Perrin: But what does it mean for a business?
Vladislav Malicevic: Essentially, that your capacity for analysis and planning now expands dramatically. Predictive analytics has always been a necessity in the business world; you can’t stay ahead of the curve if you don’t know where the curve is. Your planning now can be based on facts using heuristics, rather than gut “assumptions”. This means instead of trying to figure out what’s coming, you can start preemptively plan your response.Let’s look at some scenarios:
Scenario 1: You work for Biker’s Best, a global wholesaler of Biker’s products. You need to forecast products for the next two years across the USA. Using existing data in your Jedox cubes, you use R to provide cluster analysis and forecast demand across customer segments for different product groups over the next two years. You can see the results immediately in your Jedox Web dashboard. Then you interactively test scenarios through your dashboard which writes-back instantly into R and your Jedox cubes. The result is a comprehensive bottom-up plan for each region to reach your CEO’s overall sales targets.
Scenario 2: You work for The People’s Cinema, a national cinema company. A sequel to a hugely popular sequel is coming out, and given that it’s the summer movie season, you need to analyse the audience response to the last film in order to schedule the movie timetable (you have a limited number of copies of the film) which maximises the potential audience. You need to take into account the demographic – we can’t simply create an aggregate for each cinema, especially given that they don’t always correspond with population centres. You can, however, use multidimensional scaling in R and load this back into a Jedox cube to drive an interactive scatter plot that shows responses and ticket sales for the first movie and use a Jedox widget to visualise this over a map.
Sam Perrin: So how can I start using this?
Vladislav Malicevic: You’ll need to download and install R. But that’s essentially it. Once you’ve upgraded to Jedox 5.1, you can use R directly as a distinct transform from the Jedox ETL, just like any other native transform type. I’ve already seen innovative business applications interactively using R and Jedox and I look forward to seeing many more over the coming months.
Greetings from here in Freiburg. If you have questions about R, please post them below!