1.4.1 About the Transparency Layer

The Oracle R Enterprise transparency layer is implemented by the OREbase, OREgraphics, and OREstats packages. These Oracle R Enterprise packages contain overloaded methods of functions in the open source R base, graphics, and stats packages, respectively. The Oracle R Enterprise packages also contain Oracle R Enterprise versions of some of the open source R functions.

With the methods and functions in these packages, you can create R objects that specify data in an Oracle Database instance. When you execute an R expression that uses such an object, the method or function transparently generates a SQL query and sends it to the database. The database then executes the query and returns the results of the operation as an R object.

A database table or view is represented by an ore.frame object, which is a subclass of data.frame. Other Oracle R Enterprise classes inherit from corresponding R classes, such as ore.vector and vector. Oracle R Enterprise maps Oracle Database data types to Oracle R Enterprise classes, such as NUMBER to ore.integer. For more information on Oracle R Enterprise data types and object mappings, see "Transparency Layer Support for R Data Types and Classes".

You can use the transparency layer methods and functions to prepare database-resident data for analysis. You can then use functions in other Oracle R Enterprise packages to build and fit models and use them to score data. For large data sets, you can do the modeling and scoring using R engines embedded in Oracle Database.

See Also:

Example 1-2 Finding the Mean of the Petal Lengths by Species in R

This example illustrates the translation of an R function invocation into SQL. It uses the overloaded Oracle R Enterprise aggregate function to get the mean of the petal lengths from the IRIS_TABLE object from Example 1-2.

aggplen = aggregate(IRIS_TABLE$Petal.Length,
                    by = list(species = IRIS_TABLE$Species),
                    FUN = mean)
Listing for Example 1-2
R> aggplen = aggregate(IRIS_TABLE$Petal.Length,
                       by = list(species = IRIS_TABLE$Species),
                       FUN = mean)
R> aggplen
              species     x
setosa         setosa 1.462
versicolor versicolor 4.260
virginica   virginica 5.552

Example 1-3 SQL Equivalent of Example 1-2

This example shows the SQL equivalent of the aggregate function in Example 1-2.

SELECT "Species", AVG("Petal.Length")
GROUP BY "Species"
ORDER BY "Species";
Species     AVG("PETAL.LENGTH") 
----------- ------------------- 
setosa       1.4620000000000002 
versicolor   4.26
virginica    5.552