Lauri Koobas created SPARK-32819:
------------------------------------
Summary: Spark SQL aggregate() fails on nested string arrays
Key: SPARK-32819
URL: https://issues.apache.org/jira/browse/SPARK-32819
Project: Spark
Issue Type: Bug
Components: SQL
Affects Versions: 3.0.0
Environment: Spark 3.0.0 on Databricks (DBR 7.2).
Reporter: Lauri Koobas
The aggregate() function seems to fail if the initial state is array(array(
some string )). Seems to work if it's an INT for example.
Example that works with a simple array:
{{select aggregate(split('abcdefgh',''), array(''), (acc, x) -> array(x))}}
Example that works with nested array and INTs:
{{select aggregate(sequence(0,9), array(array(0)), (acc, x) ->
array(array(x)))}}
Example that errors:
{{select aggregate(split('abcdefgh',''), array(array('')), (acc, x) ->
array(array(x)))}}
Producing the following (shortened) error:
{{data type mismatch: argument 3 requires array<array<string>> type, however
... is of array<array<string>> type.}}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
