Practical Data Science with SAP: Machine Learning Techniques for Enterprise Data

by Greg Foss, Paul Modderman

Overview

Are you using SAP ERP and eager to unlock the enormous value of its data? With this practical guide, SAP veterans Greg Foss and Paul Modderman show you how to use several data analysis tools to solve interesting problems with your SAP data. Throughout the book, you’ll follow a fictional company as it tackles real scenarios.

Using actual data to create example code and visualizations, SAP business analysts will learn practical methods for gaining deeper insights into their business’s data. Data engineers and data scientists will explore ways to add SAP data to their analysis processes. Through grounded explanations of both SAP processes and data science tools, you’ll discover powerful methods for discovering data truths.

• Use data to tell revealing stories about your customers
• Model purchase requisition data using exploratory data analysis
• Create an anomaly detection system for SAP sales orders
• Use R and Python to make predictions on sales data
• Cluster and segment your customers based on their buying habits
• Use association rule learning to discover customer buying patterns
• Apply NLP to uncover the most highly actionable customer complaints

Product Details

ISBN-13:	9781492046448
Publisher:	O'Reilly Media, Incorporated
Publication date:	10/08/2019
Pages:	330
Product dimensions:	7.00(w) x 9.10(h) x 0.80(d)

About the Author

Greg Foss fuses battle-tested deep SAP knowledge with a passion for all things data science. His SAP career spans all areas of the technology stack - server, database, security, back and front end development, and functional expertise. As an enterprise architect, he’s been the steady guiding hand for years of managing, supporting, and enhancing SAP. As the founder of Blue Diesel Data Science, he focuses years of R, Python, machine learning algorithms, and analytics expertise on finding unique stories to tell from enterprise SAP data. Through Blue Diesel, Greg regularly contributes unique knowledge and insight into the data science blogging community, and is the principal developer and architect of VisionaryRX, an innovative pharmaceutical data dashboarding product.

Paul Modderman loves creating things and sharing them. His tech career has spanned web applications with technologies like .NET, Java, Python, and React to SAP solutions in ABAP, OData and SAPUI5, to cloud technologies in Google Cloud Platform, Amazon Web Services, and Microsoft Azure. He was principal technical architect on Mindset's certified solutions CloudSimple and Analytics for BW. He's an SAP Developer Hero, honored in 2017. Paul is the author of two books: Mindset Perspectives: SAP Development Tips, Tricks, and Projects, and the SAP Press published SAPUI5 and SAP Fiori: The Psychology of UX Design.

Table of Contents

Preface vii

1 Introduction 1

Telling Better Stories with Data 1

A Quick Look: Data Science for SAP Professionals 3

A Quick Look: SAP Basics for Data Scientists 6

Getting Data Out of SAP 8

Roles and Responsibilities 11

Summary 11

2 Data Science for SAP Professionals 13

Machine Learning 14

Supervised Machine Learning 15

Unsupervised Machine Learning 18

Semi-Supervised Machine Learning 21

Reinforcement Machine Learning 22

Neural Networks 26

Summary 43

3 SAP for Data Scientists 45

Getting Started with SAP 46

The ABAP Data Dictionary 49

Tables 50

Structures 53

Data Elements and Domains 54

Where-Used 58

ABAP QuickViewer 62

SE16 Export 68

OData Services 68

Core Data Services 80

Summary 91

4 Exploratory Data Analysis with R 93

The Four Phases of EDA 95

Phase 1 Collecting Our Data 96

Importing with R 104

Phase 2 Cleaning Our Data 107

Null Removal 107

Binary Indicators 107

Removing Extraneous Columns 108

Whitespace 108

Numbers 109

Phase 3 Analyzing Our Data 109

DataExplorer 110

Discrete Features 113

Continuous Features 117

Phase 4 Modeling Our Data 121

TensorFlow and Keras 122

Training and Testing Split 122

Shaping and One-Hot Encoding 123

Recipes 124

Preparing Data for the Neural Network 126

Results 130

Summary 132

5 Anomaly Detection with Rand Python 133

Types of Anomalies 134

Tools in R 135

AnomalyDetection 135

Anomalize 136

Getting the Data 136

SAP ECC System 137

SAP NetWeaver Gateway 142

SQL Server 153

Finding Anomalies 174

PowerBI and R 174

PowerBI and Python 184

Summary 189

6 Predictive Analytics in Rand Python 191

Predicting Sales in R 193

Step 1 Identify Data 193

Step 2 Gather Data 193

Step 3 Explore Data 194

Step 4 Model Data 195

Step 5 Evaluate Model 206

Predicting Sales in Python 210

Step 1 Identify Data 210

Step 2 Gather Data 210

Step 3 Explore Data 216

Step 4 Model Data 219

Step 5 Evaluate Model 220

Summary 222

7 Clustering and Segmentation in R 225

Understanding Clustering and Segmentation 226

RFM 227

Pareto Principle 228

k-Means 229

k-Medoid 230

Hierarchical Clustering 231

Time-Series Clustering 233

Step 1 Collecting the Data 233

Step 2 Cleaning the Data 234

Step 3 Analyzing the Data 240

Revisiting the Pareto Principle 240

Finding Optimal Clusters 241

k-Means Clustering 244

k-Medoid Clustering 249

Hierarchical Clustering 253

Manual RFM 255

Step 4 Report the Findings 258

R Markdown Code 261

R Markdown Knit 262

Summary 264

8 Association Rule Mining 267

Understanding Association Rule Mining 269

Support 269

Confidence 269

Lift 270

Apriori Algorithm 270

Operatiorealization Overview 270

Collecting the Data 271

Cleaning the Data 276

Analyzing the Data 277

Fiori 282

Summary 287

9 Natural Language Processing with the Google Cloud Natural Language API 289

Understanding Natural Language Processing 290

Sentiment Analysis 290

Translation 292

Preparing the Cloud API 292

Collecting the Data 298

Analyzing the Data 301

Summary 303

10 Conclusion 305

Original Mission 305

Recap 306

Chapter 1 Introduction 306

Chapter 2 Data Science for SAP Professionals 306

Chapter 3 SAP for Data Scientists 306

Chapter 4 Exploratory Data Analysis 307

Chapter 5 Anomaly Detection with R and Python 307

Chapter 6 Prediction with R 307

Chapter 7 Clustering and Segmentation in R 307

Chapter 8 Association Rule Mining 307

Chapter 9 Natural Language Processing with the Google Cloud Natural Language API 308

Tips and Recommendations 308

Be Creative 308

Be Practical 308

Enjoy the Ride 309

Stay in Touch 309

Index 311