The JOIN component joins tables in the map stage instead of the reduce stage. This way, transmission of a large amount of data in the shuffle stage is not required, and jobs are optimized. For example, if you need to join a large table with a small table, you can load data in the small table to the memory. This improves the operation efficiency.
Configure the component
You can configure the JOIN component only on the pipeline page of Machine Learning Designer. The following table describes the parameters.
Parameter | Description |
Join Type | The join type. Valid values: Left Join, Inner Join, Right Join, and Full Join. |
MapJoin Optimization | Specifies whether to load data in the small table to the memory to accelerate the execution of the JOIN operation. Valid values:
|
Join Condition | The join conditions, which are in the format of equations. You can manually add or remove join conditions. |
Select Output Columns from the Left Table | The output columns from the left table. |
Select Output Columns from the Right Table | The output columns from the right table. |