References → Join Recipe

The Join recipe is a recipe in which two datasets can be blended based on a set of key fields. Joins will not need to exist in the physical schema to join two datasets.

Configuration

ConfigurationDescription
Recipe NameA freeform name of how a user would like to name a recipe
Join TypeSelect one of the following join types:
  ●  Inner
  ●  Outer
  ●  Left Outer
  ●  Right Outer
  ●  Left Semi
Match OnSelect a matching criteria:

All conditions are satisfied. If multiple join clauses are created, they must ALL be true to be considered a match. This is the equivalent of setting join conditions to AND.

At least one of the conditions is satisfied. If multiple join clauses are created, only one may be true to be considered a match. This is the equivalent of setting join conditions to OR.

An Expression. Write a join condition via pySpark.
Left InputThe dataset is to be input to the left side of the join.
Right InputThe dataset is to be input to the right side of the join.
Join ConditionSelect the key columns from the right and left inputs to join on.

Join Type Descriptions

INNER JOIN Combines rows from two tables where a record exists in both tables. Outer Join: Often called a full outer join, it combines rows from two tables with NULLs assigned to the left or right dataset in places where no record relationship was present.

LEFT OUTER JOIN Combines rows from two tables, returning only matched rows from the right table; unmatched rows from the left input will have NULLS for any data from the right table.

RIGHT OUTER JOIN Combines rows from two tables, returning only matched rows from the left table; unmatched rows from the right input will have NULLS for any data from the left table.

LEFT SEMI JOIN Combines rows from two tables, returning only matched rows from the right table; unmatched rows from the left input will be dropped.

Condition Operators

  • = (Equals)
  • != (Does not equal)
  • >= (Greater than or equal to)
  • <= (Less than or equal to)
  • > (Greater than)
  • < (Less than)