named_struct

Creates a struct with the given field names and values.

Syntax

from pyspark.sql import functions as sf

sf.named_struct(*cols)

Parameters

Parameter	Type	Description
cols	pyspark.sql.Column or column name	List of columns to work on.

Returns

pyspark.sql.Column:

Examples

import pyspark.sql.functions as sf
df = spark.createDataFrame([(1, 2)], ['a', 'b'])
df.select("*", sf.named_struct(sf.lit('x'), df.a, sf.lit('y'), "b")).show()

+---+---+------------------------+
|  a|  b|named_struct(x, a, y, b)|
+---+---+------------------------+
|  1|  2|                  {1, 2}|
+---+---+------------------------+