The output of the join node is a data flow. If you perform a complex transformation such as a join, saving the result of the transformation as a table results in faster subsequent operations.
Follow these steps to create a table:
Create a Create Table or View node: In the Component Palette, click the Data section, and left-click Create Table. Move the cursor to the workflow and click again.
Right-click JoinData
, select Connect, and connect JoinData
to the create table node.
Double-click the create table node. In the Edit Create Table or View Node dialog, name the output table FAST_TABLE
.
Tell me more about what I see in the GUI.
The name of the Create Table node changes to FAST_TABLE
, the name of the output table.
FAST_TABLE
and select Run to create the table.Now use FAST_TABLE
as a data source to build classification models:
Create a Classification node: In the Component Palette, click the Models section, and
left-click Classification. Move the cursor to the workflow and click again. Name the
classification build node PredictGender
.
Right-click FAST_TABLE
, select Connect and connect the table to PredictGender
.
The Edit Classification Node dialog opens. Select CUST_GENDER
as the target. Click OK.
Right-click PredictGender
and select Run.
When the execution of PredictGender
finishes, you have built and tested four classification models. You can view
and compare the models. If you view the models, they show how they tie together the sales data (aggregated data) and the customer
demographic data as part of the analysis.
Copyright © 2011, 2012, Oracle. All rights reserved.