Contents

Title and Copyright Information

Preface

• Audience
• Documentation Accessibility
• Related Documents
• Oracle R Enterprise Online Resources
• Conventions

Changes in This Release for Oracle R Enterprise

• Changes in Oracle R Enterprise 1.5
  • New Features in Oracle R Enterprise 1.5
    • R for Datastore and Script Repository Management
    • PL/SQL and Data Dictionary Views for Datastore and Script Repository Management
    • ore.groupApply Function Changes
    • ore.randomForest Modeling Function
    • ore.summary Function Changes
    • Statistical Function Method Changes
    • Support for BLOB and CLOB Data Types
• Changes in Oracle R Enterprise 1.4.1
  • New Features in Oracle R Enterprise 1.4.1
  • Other Changes in Oracle R Enterprise 1.4.1
• Changes in Oracle R Enterprise 1.4
  • New Features in Oracle R Enterprise 1.4
• Changes in Oracle R Enterprise 1.3
  • New Features in Oracle R Enterprise 1.3
  • Other Changes in Oracle R Enterprise 1.3
• Changes in Oracle R Enterprise 1.1

1 Introducing Oracle R Enterprise

• 1.1 About Oracle R Enterprise
• 1.2 Advantages of Oracle R Enterprise
• 1.3 Get Online Help for Oracle R Enterprise Classes, Functions, and Methods
• 1.4 About Transparently Using R on Oracle Database Data
  • 1.4.1 About the Transparency Layer
  • 1.4.2 Transparency Layer Support for R Data Types and Classes
    • 1.4.2.1 About Oracle R Enterprise Data Types and Classes
    • 1.4.2.2 About the ore.frame Class
    • 1.4.2.3 Support for R Naming Conventions
    • 1.4.2.4 About Coercing R and Oracle R Enterprise Class Types
• 1.5 Typical Operations in Using Oracle R Enterprise
• 1.6 Oracle R Enterprise Global Options
• 1.7 Oracle R Enterprise Examples
  • 1.7.1 Listing the Oracle R Enterprise Examples
  • 1.7.2 Running an Oracle R Enterprise Example Script

2 Getting Started with Oracle R Enterprise

• 2.1 Connecting to an Oracle Database Instance
  • 2.1.1 About Connecting to the Database
    • 2.1.1.1 About Using the ore.connect Function
    • 2.1.1.2 About Using the ore.disconnect Function
  • 2.1.2 Using the ore.connect and ore.disconnect Functions
• 2.2 Creating and Managing R Objects in Oracle Database
  • 2.2.1 Creating R Objects for In-Database Data
    • 2.2.1.1 About Creating R Objects for Database Objects
    • 2.2.1.2 Using the ore.sync Function
    • 2.2.1.3 Using the ore.get Function
    • 2.2.1.4 Using the ore.attach Function
  • 2.2.2 Creating Ordered and Unordered ore.frame Objects
    • 2.2.2.1 About Ordering in ore.frame Objects
    • 2.2.2.2 Global Options Related to Ordering
    • 2.2.2.3 Ordering Using Keys
    • 2.2.2.4 Ordering Using Row Names
    • 2.2.2.5 Using Ordered Frames
  • 2.2.3 Moving Data to and from the Database
  • 2.2.4 Creating and Deleting Database Tables
  • 2.2.5 Saving and Managing R Objects in the Database
    • 2.2.5.1 About Persisting Oracle R Enterprise Objects
    • 2.2.5.2 About Oracle R Enterprise Datastores
    • 2.2.5.3 Saving Objects to a Datastore
    • 2.2.5.4 Control Access to Datastores
    • 2.2.5.5 Getting Information about Datastore Contents
    • 2.2.5.6 Restoring Objects from a Datastore
    • 2.2.5.7 Deleting a Datastore
    • 2.2.5.8 About Using a datastore in Embedded R Execution

3 Preparing and Exploring Data in the Database

• 3.1 Preparing Data in the Database Using Oracle R Enterprise
  • 3.1.1 About Preparing Data in the Database
  • 3.1.2 Selecting Data
    • 3.1.2.1 Selecting Data by Column
    • 3.1.2.2 Selecting Data by Row
    • 3.1.2.3 Selecting Data by Value
  • 3.1.3 Indexing Data
  • 3.1.4 Combining Data
  • 3.1.5 Summarizing Data
  • 3.1.6 Transforming Data
  • 3.1.7 Sampling Data
  • 3.1.8 Partitioning Data
  • 3.1.9 Preparing Time Series Data
• 3.2 Exploring Data
  • 3.2.1 About the Exploratory Data Analysis Functions
  • 3.2.2 About the NARROW Data Set for Examples
  • 3.2.3 Correlating Data
  • 3.2.4 Cross-Tabulating Data
  • 3.2.5 Analyzing the Frequency of Cross-Tabulations
  • 3.2.6 Building Exponential Smoothing Models on Time Series Data
  • 3.2.7 Ranking Data
  • 3.2.8 Sorting Data
  • 3.2.9 Summarizing Data with ore.summary
  • 3.2.10 Analyzing Distribution of Numeric Variables
  • 3.2.11 Principal Component Analysis
  • 3.2.12 Singular Value Decomposition
• 3.3 Using a Third-Party Package on the Client

4 Building Models in Oracle R Enterprise

• 4.1 Building Oracle R Enterprise Models
  • 4.1.1 About OREmodels Functions
  • 4.1.2 About the longley Data Set for Examples
  • 4.1.3 Building Linear Regression Models
  • 4.1.4 Building a Generalized Linear Model
  • 4.1.5 Building a Neural Network Model
  • 4.1.6 Building a Random Forest Model
• 4.2 Building Oracle Data Mining Models
  • 4.2.1 About Building Oracle Data Mining Models using Oracle R Enterprise
    • 4.2.1.1 Oracle Data Mining Models Supported by Oracle R Enterprise
    • 4.2.1.2 About Oracle Data Mining Models Built by Oracle R Enterprise Functions
  • 4.2.2 Building an Association Rules Model
  • 4.2.3 Building an Attribute Importance Model
  • 4.2.4 Building a Decision Tree Model
  • 4.2.5 Building General Linearized Models
  • 4.2.6 Building a k-Means Model
  • 4.2.7 Building a Naive Bayes Model
  • 4.2.8 Building a Non-Negative Matrix Factorization Model
  • 4.2.9 Building an Orthogonal Partitioning Cluster Model
  • 4.2.10 Building a Support Vector Machine Model
• 4.3 Cross-Validating Models

5 Predicting With R Models

• 5.1 About the ore.predict Function
• 5.2 Using the ore.predict Function

6 Using Oracle R Enterprise Embedded R Execution

• 6.1 About Oracle R Enterprise Embedded R Execution
  • 6.1.1 Benefits of Embedded R Execution
  • 6.1.2 APIs for Embedded R Execution
  • 6.1.3 Security Considerations for Scripts
  • 6.1.4 Support for Parallel Execution
  • 6.1.5 Installing a Third-Party Package for Use in Embedded R Execution
• 6.2 R Interface for Embedded R Execution
  • 6.2.1 Arguments for Functions that Run Scripts
    • 6.2.1.1 Input Function to Execute
    • 6.2.1.2 Optional and Control Arguments
    • 6.2.1.3 Structure of Return Value
    • 6.2.1.4 Input Data
    • 6.2.1.5 Parallel Execution
    • 6.2.1.6 Unique Arguments
  • 6.2.2 Manage Scripts in R
  • 6.2.3 Using the ore.doEval Function
  • 6.2.4 Using the ore.tableApply Function
  • 6.2.5 Using the ore.groupApply Function
    • 6.2.5.1 Partitioning on a Single Column
    • 6.2.5.2 Partitioning on Multiple Columns
  • 6.2.6 Using the ore.rowApply Function
  • 6.2.7 Using the ore.indexApply Function
    • 6.2.7.1 Simple Example of Using the ore.indexApply Function
    • 6.2.7.2 Column-Parallel Use Case
    • 6.2.7.3 Simulations Use Case
• 6.3 SQL Interface for Embedded R Execution
  • 6.3.1 About Oracle R Enterprise SQL Table Functions
    • 6.3.1.1 Parameters of the SQL Table Functions
    • 6.3.1.2 Return Value of SQL Table Functions
    • 6.3.1.3 Connecting to Oracle R Enterprise in Embedded R Execution
  • 6.3.2 Manage Scripts in SQL
    • 6.3.2.1 Creating a Script with the SQL APIs
  • 6.3.3 Manage Datastores in SQL

A SQL APIs for Oracle R Enterprise

• A.1 rqDropDataStore Procedure
• A.2 rqEval Function
• A.3 rqGrant Procedure
• A.4 rqGroupEval Function
• A.5 rqRevoke Procedure
• A.6 rqRowEval Function
• A.7 rqTableEval Function
• A.8 sys.rqScriptCreate Procedure
• A.9 sys.rqScriptDrop Procedure

B Oracle Database Views for Oracle R Enterprise

• B.1 ALL_RQ_DATASTORES
• B.2 ALL_RQ_SCRIPTS
• B.3 RQUSER_DATASTORECONTENTS
• B.4 RQUSER_DATASTORELIST
• B.5 USER_RQ_DATASTORE_PRIVS
• B.6 USER_RQ_DATASTORES
• B.7 USER_RQ_SCRIPT_PRIVS
• B.8 USER_RQ_SCRIPTS

C R Operators and Functions Supported by Oracle R Enterprise

Index