SAP Hana Advanced Data Modeling

Anil Babu Ankisettipalli is a post graduate in computer applications and has more than 17 years of experience in developing application and technology software. Since 2010, he has designed and developed SAP HANA applications and products using predictive analytics, machine learning, and data mining algorithms. At SAP Labs, he uses data science and big data technologies within the SAP HANA platform to solve critical problems of SAP customers. Hansen Chen is an experienced database expert and graduate of Peking University in China. In 2006 he joined Business Objects, where he worked on Web Intelligence and the semantic layer. He joined the SAP Strategic Customer Engagements group as a senior specialist in 2011, and worked with customers in co-innovation projects on top of the SAP HANA platform. His specific area of SAP HANA focus is on modeling and performance tuning. Pranav Wankawala has more than 13 years of experience in software engineering and developing enterprise applications. He started working on SAP HANA in 2009, when it was still a research project, and designed and architected one of the first SAP products that ran on the platform. During his tenure at SAP, he lead a team of highly skilled engineers and technologists that helped SAP customers improve their day-to-day business with SAP HANA. Today, Pranav is a director of engineering at PredictSpring, a company that is revolutionizing the way mobile commerce works.

1 … SAP HANA Data Models ... 21

1.1 … SAP HANA Database Architecture Overview ... 21
1.2 … SAP HANA Modeling Paradigms ... 22
1.3 … Information Views ... 26
1.4 ... Analytic Privileges ... 67
1.5 … Stored Procedures ... 75
1.6 … Application Function Library ... 86
1.7 … Summary ... 90

2 … Modeling Complex Logic ... 93

2.1 … Achieving Recursive Logic with Hierarchies ... 93
2.2 … Transposing Columns and Rows ... 110
2.3 … Using cube() with Hierarchies ... 123
2.4 … Calculating Running Total ... 127
2.5 … Calculating Cumulative Sum ... 131
2.6 … Filtering Data Based on Ranking ... 134
2.7 … Controlling Join Paths via Filters ... 138
2.8 … Full Outer Join in a Calculation View ... 143
2.9 … Making Dynamic Queries in a Stored Procedure ... 148
2.10 … Showing History Records Side By Side ... 153
2.11 … Sample Data ... 158
2.12 … Using a Vertical Union to Join Tables ... 161
2.13 … Sorting Records ... 163
2.14 … Finding Missing Values ... 168
2.15 … Using Window Functions for Complex Grouping ... 172
2.16 … Joining Based on a Date Sequence ... 178
2.17 … Using a Nested Calculation View ... 185
2.18 … Summary ... 191

3 … Scaling for Large Datasets ... 193

3.1 … Partitioning ... 193
3.2 … Using Input Parameters to Enforce Pruning ... 198
3.3 … Creating an Index ... 201
3.4 … Analyzing Query Performance with Tools ... 205
3.5 … Enforcing Execution Paths ... 214
3.6 … Using a Union with Constant Values Instead of a Join ... 218
3.7 … Manipulating Joins in an Analytic View ... 224
3.8 … Time Traveling ... 236
3.9 … Storing Temporary Data ... 243
3.10 … Calculating Count Distinct ... 247
3.11 … Using Cached Views ... 250
3.12 … Summary ... 258

4 … Basic Predictive Modeling ... 259

4.1 … Predictive Analytics Lifecycle in SAP HANA ... 259
4.2 … Data Exploration ... 270
4.3 … Data Preparation ... 291
4.4 … Modeling ... 297
4.5 … Creating Models Using SAP Applications on SAP HANA ... 308
4.6 … Summary ... 318

5 … Advanced Predictive Modeling ... 319

5.1 … R Script Modeling and Design ... 319
5.2 … PAL Model Consumption ... 326
5.3 … Real-Time Model Consumption vs. Batch Predictive Modeling ... 329
5.4 … Impact of Data Partitions in Predictive Modeling ... 332
5.5 … Using Multiple R Servers and Data Partitions ... 333
5.6 … Modeling Using R and PAL Simultaneously ... 337
5.7 … Summary ... 340

6 … Simulations and Optimizations ... 341

6.1 … Case Study ... 341
6.2 … Monte Carlo Simulation of Value-at-Risk ... 342
6.3 … Portfolio Optimization … 363
6.4 … Summary ... 380

The Authors ... 381