Kontext Kontext / Code Snippets & Tips

PySpark DataFrame - union, unionAll and unionByName

event 2022-08-16 visibility 2,153 comment 0 insights
more_vert
insights Stats

Code description

PySpark DataFrame provides three methods to union data together: union, unionAll and unionByName. The first two are like Spark SQL  UNION ALL clause which doesn't remove duplicates.  unionAll is the alias for union. We can use distinct method to deduplicate. The third function will use column names to resolve columns instead of positions. unionByName can also be used to merge two DataFrames with different schemas. 

Syntax of unionByName

DataFrame.unionByName(other, allowMissingColumns=False)

If allowMissingColumns is specified as True, the missing columns in both DataFrame will be added with default value null. This parameter is only available from Spark 3.1.0. For lower versions, the follow error may appear:

    df1.unionByName(df4, allowMissingColumns=True).show(truncate=False)
TypeError: unionByName() got an unexpected keyword argument 'allowMissingColumns'

Outputs

The following are the outputs from the code snippet.

+---+---+---+
|c1 |c2 |c3 |
+---+---+---+
|1  |A  |100|
|2  |B  |200|
+---+---+---+

+---+---+---+
|c1 |c2 |c3 |
+---+---+---+
|1  |A  |100|
|1  |A  |100|
|2  |B  |200|
+---+---+---+

+---+---+---+
|c1 |c2 |c3 |
+---+---+---+
|1  |A  |100|
|2  |B  |200|
+---+---+---+

+---+---+----+----+
|c1 |c2 |c3  |c4  |
+---+---+----+----+
|1  |A  |100 |null|
|3  |C  |null|ABC |
+---+---+----+----+

Code snippet

from pyspark.sql import SparkSession

appName = "PySpark union Examples"
master = "local"

spark = SparkSession.builder \
    .appName(appName) \
    .master(master) \
    .getOrCreate()

spark.sparkContext.setLogLevel("ERROR")

# Create DataFrames
df1 = spark.createDataFrame([[1, 'A', 100]], ['c1', 'c2', 'c3'])
df2 = spark.createDataFrame([[1, 'A', 100], [2, 'B', 200]], ['c1', 'c2', 'c3'])
df3 = spark.createDataFrame([[2, 200, 'B']], ['c1',  'c3', 'c2'])
df4 = spark.createDataFrame([[3, 'C', 'ABC']], ['c1', 'c2', 'c4'])

# Use union method
df1.union(df2).distinct().show(truncate=False)

# Use unionAll method
df1.unionAll(df2).show(truncate=False)

# Use unionByName method
df1.unionByName(df3).show(truncate=False)
df1.unionByName(df4, allowMissingColumns=True).show(truncate=False)
More from Kontext
comment Comments
No comments yet.

Please log in or register to comment.

account_circle Log in person_add Register

Log in with external accounts