pyspark.sql.functions.try_divide¶

pyspark.sql.functions.try_divide(left: ColumnOrName, right: ColumnOrName) → pyspark.sql.column.Column[source]¶

Returns dividend/divisor. It always performs floating point division. Its result is always null if divisor is 0.

New in version 3.5.0.

Parameters

leftColumn or str: dividend
rightColumn or str: divisor

Examples

>>> df = spark.createDataFrame([(6000, 15), (1990, 2)], ["a", "b"])
>>> df.select(try_divide(df.a, df.b).alias('r')).collect()
[Row(r=400.0), Row(r=995.0)]

>>> df = spark.createDataFrame([(1, 2)], ["year", "month"])
>>> df.select(
...     try_divide(make_interval(df.year), df.month).alias('r')
... ).show(truncate=False)
+--------+
|r       |
+--------+
|6 months|
+--------+

>>> df.select(
...     try_divide(make_interval(df.year, df.month), lit(2)).alias('r')
... ).show(truncate=False)
+--------+
|r       |
+--------+
|7 months|
+--------+

>>> df.select(
...     try_divide(make_interval(df.year, df.month), lit(0)).alias('r')
... ).show(truncate=False)
+----+
|r   |
+----+
|NULL|
+----+

pyspark.sql.functions.try_avg

pyspark.sql.functions.try_multiply