[
https://issues.apache.org/jira/browse/ARROW-13436?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17391743#comment-17391743
]
Joris Van den Bossche commented on ARROW-13436:
-----------------------------------------------
Yes, I think an empty list should mean no columns read in (which is the current
behaviour?). Note that the table still has the correct length ({{num_rows}}),
even though it has no rows. Would be good to clarify in the docs.
> [Python][Doc] Clarify what should be expected if read_table is passed an
> empty list of columns
> ----------------------------------------------------------------------------------------------
>
> Key: ARROW-13436
> URL: https://issues.apache.org/jira/browse/ARROW-13436
> Project: Apache Arrow
> Issue Type: Bug
> Components: Python
> Reporter: Weston Pace
> Assignee: Weston Pace
> Priority: Major
>
> The documentation for pyarrow.parquet.read_table states:
>
> * *columns* (_list_) – If not None, only these columns will be read from the
> file. A column name may be a prefix of a nested field, e.g. ‘a’ will select
> ‘a.b’, ‘a.c’, and ‘a.d.e’.
>
> It is not clear what should be the expected result if columns is an empty
> list. In pyarrow 3.0 this read in all columns (as long as
> use_legacy_dataset=False). In pyarrow 4.0 this doesn't read in any columns.
> I think this behavior (not reading in any columns) is the correct behavior
> (since None can be used for all columns) but we should clarify that in the
> docs.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
