Oracle Data Warehousing and Business Intelligence - Oracle Essentials - page 258

Processing. Tools that can work against either relational databases or MOLAP engines
are sometimes referred to as hybrid tools .
The Oracle OLAP Option can be thought of as a MOLAP cube within the relational
database, accessible via SQL. Though available since Oracle 9 i , Oracle Database 11 g
significantly improved the flexibility of accessing the OLAP Option. Business users
formerly needed to specifically point their queries to OLAP Option cubes. Since Oracle
Database 11 g , the OLAP cubes can be used transparently as an alternative to materialized
views because Oracle's SQL query rewrite recognizes the cubes. The materialized view
refresh can refresh OLAP cubes as of Oracle Database 11 g .
OLAP Option cubes are deployed in what are called analytic workspaces . They can be
created using a simplified logical dimensional modeling tool called the Analytic Work‐
space Manager (AWM). The tool provides an interface for creation of the cubes and for
building maps from relational tables into the cubes.
Custom OLAP applications can be built using Oracle's JDeveloper and business intelli‐
gence beans , although this is much less common than using off-the-shelf tools. The Java
beans provide prebuilt components for manipulating tables, crosstabs, and graphs, and
for building queries and calculations similar to the functionality previously found in
Express. JDeveloper generates Java code utilizing these building blocks that maps to the
Java OLAP API provided by Oracle's OLAP Option.
Analytics and Statistics in the Database
Analysis of large data sets is faster when it takes place where the data is stored, since this
approach avoids the overhead of moving the data sets around. This section describes
the database functions and other features available for analytics, statistics, and data
mining in the Oracle Database.
It is worth noting here that the growing use of Oracle for statistical computations led
to support for floating-point number types providing the precision outlined in the IEEE
754-1985 standard (with minor differences). These are provided in the datatypes BI‐
NARY_FLOAT and BINARY_DOUBLE in Oracle Database 10 g and more recent da‐
tabase releases.
Basic Analytic and Statistical Functions
Oracle releases dating back to Oracle8 i have continued to add new analytic and statistical
functions as SQL extensions to the core Oracle Enterprise Edition and Standard Edition
databases. These analytic functions now include: