SAP Blog
SAP Predictive Analysis Installation

SAP Predictive Analysis Installation

SAP Predictive Analysis is the latest addition to the SAP BusinessObjects suite and introduces entirely new functionality to the existing Business Objects toolbox.  Predictive Analysis integrates SAP’s Visual Intelligence data visualization tool with new predictive functionality powered by both open source R and SAP-written algorithms.  Predictive Analysis includes algorithms for time series forecasting (for predicting sales, demand, price, and other time-dependent metrics), clustering (for identifying distinct groups of individuals based on numeric descriptive data), decision trees (for creating a tree-like set of decision support rules to categorize observations), and linear regression (for fitting linear relationships between a dependent variable and one or more predictors).  These predictive algorithms can be used to extract insights and predictions, improving the value and actionability of the existing Business Intelligence infrastructure.  Predictive Analysis combines these powerful predictive algorithms with a familiar and easy-to-use tool that integrates with existing BusinessObjects tools to make data preparation, model building, and implementation faster and easier than ever before. 
SAP Predictive Analysis is installed locally on the user’s machine and accesses data for processing locally (from a CSV, Excel, or ODBC connection to a database) or on SAP HANA.  Predictive Analysis can be used alone to analyze data on the client machine or can be paired with HANA, allowing PA to leverage the powerful in-memory processing power of SAP HANA. 

Today’s blog post is a brief how-to for getting SAP Predictive Analysis installed and working for both local data and on HANA.  Since Predictive Analysis has the power to run local R algorithms, HANA-R algorithms, and HANA Predictive Analysis Library (PAL) algorithms, installation is somewhat complex.  Installation steps include:

1.)    Installing PA
2.)    Installing R on local machine
3.)    PA HANA PAL installation/activation
4.)    R and Rserve Linux Installation and HANA Rserve setup

The following items are required to perform the full installation:

•    Predictive Analysis installation files for version 1.0.8 (or latest available)
•    An internet connection to install R and required packages
•    SAP HANA
•    A Linux R/Rserve host (SAP currently only supports SUSE Linux)

Download full PDF and Continue Reading…

Hilary BlissAbout Hillary Bliss
Hillary Bliss is a Business Intelligence Consultant specializing in data warehouse design, ETL development, statistical analysis, and predictive modeling.  Hillary works with clients and vendors to integrate business analysis and predictive modeling solutions into data warehouses based on their data and business needs. With Decision First Technologies, Hillary uses Data Services, Web Intelligence, Predictive Analysis, and HANA.  Hillary has a Master’s in Statistics and an MBA from Georgia Tech.

protiviti

8 comments

  • If i had Business Objects, i can install Predictive Analysis? or i must purchase Predictive Analysis?
    Thanks

  • Hillary,

    Thanks for the detailed blog with screenshots. I am working with large datasets (31 M rows, 20 GB data and will be running segmentation and association models on that. I have a fairly large server for R ( 2*8 cores, 128 GB RAM) residing on HANA and will be using that for modelling.

    If I use a script and integrate that with SAP PA , which is installed on my 4GB laptop, and use R code, where will the memory and CPU be taken from. I don’t have any hopes of getting those models run successfully from my machine and would need to use the R server memory. If that is the case, what if anything will run on my machine.

    Regards,
    Sujoy Roychowdhury
    Lead, Advanced Analytics CoE
    Ericsson AB
    India

    • Sujoy—

      Firstly, with that volume of data you definitely want to be using PA in HANA Online mode, which ensures that the data remains on the HANA server and does not use any resources from your laptop. With high volumes of data I would recommend using the HANA PAL algorithms in PA vs. the HANA R algorithms. With significant volumes of transactional data (several million records representing ~1.5 million transactions), my experience was that the PAL Apriori association algorithm ran in much less than 10 seconds.

      Hillary

      • Thanks Hilary for the quick response. However will I not be restricted algorithwise if I use HANA PAL algorithms. For example, I can code an incremental mode of processing or define distances and measures in a segmentation model with more flexibility on R than HANA PAL (as far as I understand). Also if I need to use R where will the processing take place – R server, HANA server or my laptop. Needless to say I will use it in online model.

        Regards,
        Sujoy

        • PAL algorithms offer the best performance on SAP HANA, and the algorithms you mentioned (segmentation and association) are supported within the PAL. The HANA-R algorithms are best when you need custom functionality. If your PA document is in HANA Online mode, the R functions will process on a separate Linux server with an Rserve instance that must be configured to work with the HANA server acting as the R client. It is not possible to run R algorithms on your local laptop instance of R when in HANA Online mode.

          Hillary