array_compact

Removes null values from the array.

Syntax

from pyspark.sql import functions as sf

sf.array_compact(col)

Parameters

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

Returns

pyspark.sql.Column: A new column that is an array excluding the null values from the input column.

Examples

Example 1: Removing null values from a simple array

from pyspark.sql import functions as sf
df = spark.createDataFrame([([1, None, 2, 3],)], ['data'])
df.select(sf.array_compact(df.data)).show()

+-------------------+
|array_compact(data)|
+-------------------+
|          [1, 2, 3]|
+-------------------+

Example 2: Removing null values from multiple arrays

from pyspark.sql import functions as sf
df = spark.createDataFrame([([1, None, 2, 3],), ([4, 5, None, 4],)], ['data'])
df.select(sf.array_compact(df.data)).show()

+-------------------+
|array_compact(data)|
+-------------------+
|          [1, 2, 3]|
|          [4, 5, 4]|
+-------------------+

Example 3: Removing null values from an array with all null values

from pyspark.sql import functions as sf
from pyspark.sql.types import ArrayType, StringType, StructField, StructType schema = StructType([StructField("data", ArrayType(StringType()), True)])
df = spark.createDataFrame([([None, None, None],)], schema)
df.select(sf.array_compact(df.data)).show()

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

Example 4: Removing null values from an array with no null values

from pyspark.sql import functions as sf
df = spark.createDataFrame([([1, 2, 3],)], ['data'])
df.select(sf.array_compact(df.data)).show()

+-------------------+
|array_compact(data)|
+-------------------+
|          [1, 2, 3]|
+-------------------+

Example 5: Removing null values from an empty array

from pyspark.sql import functions as sf
from pyspark.sql.types import ArrayType, StringType, StructField, StructType
schema = StructType([
  StructField("data", ArrayType(StringType()), True)
])
df = spark.createDataFrame([([],)], schema)
df.select(sf.array_compact(df.data)).show()

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