To begin using Oracle R Enterprise, you first connect to a schema in an Oracle Database instance with the ore.connect
function. Only one Oracle R Enterprise connection can exist at a time during an R session. If an R session is already connected to the database, then invoking ore.connect
terminates the active connection before opening a new connection. Before attempting to connect, you can discover whether an active connection exists by using the ore.is.connected
function.
You explicitly end a connection with the ore.disconnect
function. If you do not invoke ore.disconnect
, then the connection is automatically terminated when the R session ends. For more information on ore.disconnect
, see "About Using the ore.disconnect Function."
With the type
argument of ore.connect
, you specify the type of connection, either ORACLE or HIVE. A HIVE type of connection connects to Hive tables in a Hadoop cluster. An ORACLE type of connection connects to a schema in an Oracle Database instance. The default value of type
is "ORACLE"
.
If the connection type is HIVE, then ore.connect
ignores all other arguments. For information on Oracle R Connector for Hadoop and Hive, see Oracle Big Data Connectors User's Guide. The HIVE option applies only if you are using Oracle R Advanced Analytics for Hadoop (ORAAH) in conjunction with a Hadoop cluster. ORAAH is part of the Oracle Big Data Connectors option to the Big Data Appliance.
If the connection type is ORACLE, then you do the following:
Use the logical all
argument to specify whether Oracle R Enterprise automatically creates an ore.frame
object for each table to which the user has access in the schema and makes those ore.frame
objects visible in the current R session. The ore.frame
objects contain metadata about the tables. The default value of the all
argument is FALSE
.
If all = TRUE
, then Oracle R Enterprise implicitly invokes the ore.sync
and ore.attach
functions. If all = FALSE
, then the user must explicitly invoke ore.sync
to create ore.frame
objects. To access these objects by name, the user must invoke ore.attach
to include the names in the search path. For information on those functions, see "Creating R Objects for In-Database Data".
Use either the conn_string
argument, or various combinations of the user
, sid
, host
, password
, port
, service_name
, and conn_string
arguments to specify information that identifies the connection.
To avoid using a clear-text password, you can specify an Oracle wallet password with the conn_string
argument. No other arguments are needed. By specifying an Oracle wallet password, you can avoid embedding a database user password in application code, batch jobs, or scripts. For information on creating an Oracle wallet, see Oracle R Enterprise Installation and Administration Guide.
With the other connection identifier arguments, you specify a database user name, host name, and password, and either a system identifier (SID) or service name, and, optionally, a TCP port, or you specify a database user name, password, and a conn_string
argument.
The default value of the port
argument is 1521, the default value of host
is "localhost"
, which specifies the local host, and the default value of conn_string
is NULL. You specify the local host when your R session is running on the same computer as the Oracle Database instance to which you want to connect.
See Also:
"Using the ore.connect and ore.disconnect Functions" for examples of using the various connection identifiers