[
https://issues.apache.org/jira/browse/ARROW-6719?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16939159#comment-16939159
]
V Luong commented on ARROW-6719:
--------------------------------
Sorry, I've just found out that the PyArrow used was actually 0.14.0.RAY,
packaged together with the Ray project. The error disappears when I just import
PyArrow 0.14.0 or 0.14.1 without importing Ray. It still begs the question of
what's the difference between PyArrow 0.14.0 and 0.14.0.RAY, but I guess
strictly this is an issue to be filed with Ray instead.
I've closed this issue here.
Thanks [~jorisvandenbossche]
> Parquet read_table error in Python3.7: pyarrow.lib.ArrowInvalid: Column data
> for field with type list<...> is inconsistent with schema list<...>
> ------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: ARROW-6719
> URL: https://issues.apache.org/jira/browse/ARROW-6719
> Project: Apache Arrow
> Issue Type: Bug
> Affects Versions: 0.14.1
> Environment: Python 3.7
> Reporter: V Luong
> Priority: Critical
> Labels: parquet
>
> I have Parquet files with certain complex columns of type List<item: double>,
> List<item: string>, etc. and am using latest PyArrow (0.14.1) to process them.
> In Python 2.7, pyarrow.parquet.read_table(...) processes these files
> correctly, without any problem.
> But in Python 3.7, the same pyarrow.parquet.read_table(...) function calls
> return errors of the following kind:
> "pyarrow.lib.ArrowInvalid: Column data for field 0 with type list<item:
> double> is inconsistent with schema list<element: double>"
> This issue might be related to
> https://issues.apache.org/jira/browse/ARROW-6068
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
