February 19, 2015 1 minute read

Extending SAP Predictive Analytics Functionality

SAP Predictive Analytics incorporates two different approaches for carrying out predictive modelling

Automated Analytics: A highly automated framework that enables non-Data-Scientists to quickly create powerful models without deep statistical expertise. This functionality was formerly known as SAP InfiniteInsight.
Expert Analytics: A graphical workbench for expert users to design specific analytical workflows. This functionality was formerly known as SAP Predictive Analysis.

This website gives an overview on how to extend the “Expert Analytics” functionality. Power Users can extend and customise the functionality of the *Expert Analytics”-mode by adding their own R-Scripts. The logic is encapsulated in so called “”R Extensions” (or formerly called “Custom R Components”), which casual users can easily work with without having to have any R skills.

This page is pointing to a number of documents that help create such components. It is also linking to various components that have been published on the web and that can be quickly implemented.

Please note however, that the code on these pages is not supported by SAP and that it is provided without any warranty.

Custom R Components	Developer Documentation
Data Understanding Measures of Location: Mean, Standard Deviation, … Correlation Matrix Density Plots for Multiple Variables Normal Distribution Test Chi Squared Test of Independence Data Preparation Add Missing Dates Aggregation Filter Missing Values Graphical Row Filter Identify Duplicate Column Values Numerical Transform Random Split Stratified Sampling Clustering Hierarchical Clustering Classification Random Forest Classification Naive Bayes Regression Logistic Regression And another Logistic Regression Forecasting Add Date Attributes Bulk Forecasting by Month 2.0 Forecast Daily Time Series Visualisation Interactive 3D Plot Confusion Matrix Google Maps Pareto Plot Advanced Graphics Miscellaneous String Generator (Random and Unique) Travelling Time and Distance Multi-Input Component	Official Documentation See the Chapter “Adding Custom Component” in the user guide. Official Product Tutorials Create a Custom R Component Video Creating New Functionality Adding further data mining algorithms & visualizations (incl.video) Using Custom R functions in SAP Predictive Analysis Modifying Custom R Component from Client-Side to Server-Side Execution Deploying New Functionality Beginning with SAP Predictive Analytics 2.2 Custom R Components can be exported and shared as .spar file.

Data Understanding

Data Preparation

Clustering

Hierarchical Clustering

Classification

Regression

Logistic Regression
And another Logistic Regression

Forecasting

Visualisation

Miscellaneous

Official Documentation

See the Chapter “Adding Custom Component” in the user guide.

Official Product Tutorials

Create a Custom R Component

Video

Creating New Functionality

Deploying New Functionality

Beginning with SAP Predictive Analytics 2.2 Custom R Components can be exported and shared as .spar file.

4 Comments

You must be Logged on to comment or reply to a post.

Former Member

July 21, 2015 at 2:23 pm

Excellent info! Thank you for sharing as always, Andreas. 🙂

Tracy

Abdel DADOUCHE (SAP)

July 22, 2015 at 8:08 am

Thank you for sharing!

Former Member

November 4, 2015 at 10:19 pm

Great source of #predictive info.
Everyone should bookmark it!

Tae-Gon Kwon

September 2, 2016 at 4:17 am

Hi Andreas,

At first glance, I thought that PA would be far behind SPSS/SAS experience.

especially SPSS have given me wonderful convenience and integration for several years.

These kinds of customization and enhancement were done in R studio, accessing relevant R library.

Now I could have a change to utilize them directly in SAP PA solution, Awesome !

- sample dataset also helps me intuitive understanding.

Every topics are well guided and balanced from business and statistics perspectives.

Thanks for valuable collections. Ted