[
https://issues.apache.org/jira/browse/SPARK-61522?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hyukjin Kwon updated SPARK-61522:
---------------------------------
Fix Version/s: connect-rust-4.3.0
> Name a Python class as PyArrow does in the not-a-sequence error of
> createDataFrame, in the Python client
> --------------------------------------------------------------------------------------------------------
>
> Key: SPARK-61522
> URL: https://issues.apache.org/jira/browse/SPARK-61522
> Project: Spark
> Issue Type: Bug
> Components: Connect
> Affects Versions: connect-rust-4.3.0
> Reporter: Hyukjin Kwon
> Priority: Major
> Fix For: connect-rust-4.3.0
>
>
> Follow-up of SPARK-61483. For a value of a Python-defined class where an
> array is expected, createDataFrame raises the error "... was not a sequence
> or recognized null for conversion to list type" naming the class with its
> module (__main__.DictSub, collections.Counter), where PyArrow names it by its
> tp_name, the bare class name (DictSub, Counter). Also,
> test_freed_instances_release_their_dict_as_pyspark asserts vars(F.col("a"))
> == {"tag": 1} as PySpark's, but PySpark's vars() of a Column also holds _expr.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]