map_entries

Returns an unordered array of all entries in the given map.

Syntax

from pyspark.sql import functions as sf

sf.map_entries(col)

Parameters

Parameter	Type	Description
col	pyspark.sql.Column or str	Name of column or expression

Returns

pyspark.sql.Column: An array of key value pairs as a struct type

Examples

Example 1: Extracting entries from a simple map

from pyspark.sql import functions as sf
df = spark.sql("SELECT map(1, 'a', 2, 'b') as data")
df.select(sf.sort_array(sf.map_entries("data"))).show()

+-----------------------------------+
|sort_array(map_entries(data), true)|
+-----------------------------------+
|                   [{1, a}, {2, b}]|
+-----------------------------------+

Example 2: Extracting entries from a map with complex keys and values

from pyspark.sql import functions as sf
df = spark.sql("SELECT map(array(1, 2), array('a', 'b'), "
  "array(3, 4), array('c', 'd')) as data")
df.select(sf.sort_array(sf.map_entries("data"))).show(truncate=False)

+------------------------------------+
|sort_array(map_entries(data), true) |
+------------------------------------+
|[{[1, 2], [a, b]}, {[3, 4], [c, d]}]|
+------------------------------------+

Example 3: Extracting entries from an empty map

from pyspark.sql import functions as sf
df = spark.sql("SELECT map() as data")
df.select(sf.map_entries("data")).show()

+-----------------+
|map_entries(data)|
+-----------------+
|               []|
+-----------------+