Problem: How to Convert StructType (struct) DataFrame Column to Map (MapType) Column which is similar to Python Dictionary (Dict).
Solution: PySpark provides a create_map() function that takes a list of column types as an argument and returns a MapType column, so we can use this to convert the DataFrame struct column to map Type. struct is a type of StructType and MapType is used to store Dictionary key-value pair.
from pyspark.sql import SparkSession
from pyspark.sql.types import StructType,StructField, StringType, IntegerType
spark = SparkSession.builder.appName('SparkByExamples.com').getOrCreate()
data = [ ("36636","Finance",(3000,"USA")),
("40288","Finance",(5000,"IND")),
("42114","Sales",(3900,"USA")),
("39192","Marketing",(2500,"CAN")),
("34534","Sales",(6500,"USA")) ]
schema = StructType([
StructField('id', StringType(), True),
StructField('dept', StringType(), True),
StructField('properties', StructType([
StructField('salary', IntegerType(), True),
StructField('location', StringType(), True)
]))
])
df = spark.createDataFrame(data=data,schema=schema)
df.printSchema()
df.show(truncate=False)
This yields the below output. Here we have a properties struct column that has 2 columns salary and location.
root
|-- id: string (nullable = true)
|-- dept: string (nullable = true)
|-- properties: struct (nullable = true)
| |-- salary: integer (nullable = true)
| |-- location: string (nullable = true)
+-----+---------+-----------+
|id |dept |properties |
+-----+---------+-----------+
|36636|Finance |[3000, USA]|
|40288|Finance |[5000, IND]|
|42114|Sales |[3900, USA]|
|39192|Marketing|[2500, CAN]|
|34534|Sales |[6500, USA]|
+-----+---------+-----------+
Convert StructType to MapType (map) Column
create_map() is a PySpark SQL function that is used to convert StructType to MapType column.
#Convert struct type to Map
from pyspark.sql.functions import col,lit,create_map
df = df.withColumn("propertiesMap",create_map(
lit("salary"),col("properties.salary"),
lit("location"),col("properties.location")
)).drop("properties")
df.printSchema()
df.show(truncate=False)
This yields below output, properties struct column has been converted to propertiesMap which is MapType(map) column.
root
|-- id: string (nullable = true)
|-- dept: string (nullable = true)
|-- propertiesMap: map (nullable = false)
| |-- key: string
| |-- value: string (valueContainsNull = true)
+-----+---------+---------------------------------+
|id |dept |propertiesMap |
+-----+---------+---------------------------------+
|36636|Finance |[salary -> 3000, location -> USA]|
|40288|Finance |[salary -> 5000, location -> IND]|
|42114|Sales |[salary -> 3900, location -> USA]|
|39192|Marketing|[salary -> 2500, location -> CAN]|
|34534|Sales |[salary -> 6500, location -> USA]|
+-----+---------+---------------------------------+
You can also achieve this programmatically with out specifying struct column name individually, but I will cover this later.
Happy Learning !!