[
https://issues.apache.org/jira/browse/ARROW-17464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17597221#comment-17597221
]
Rok Mihevc commented on ARROW-17464:
------------------------------------
I'm not sure if this is the best forum for this question, but should we
consider bfloat16 (and perhaps other non-IEEE defined types)? See
https://en.wikipedia.org/wiki/Bfloat16_floating-point_format
> [C++] Support float16 in writing/reading parquet
> ------------------------------------------------
>
> Key: ARROW-17464
> URL: https://issues.apache.org/jira/browse/ARROW-17464
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++, Parquet, Python
> Reporter: Anja Boskovic
> Assignee: Anja Boskovic
> Priority: Major
> Labels: parquet, pull-request-available
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Half-float values are not supported in Parquet. Here is a previous issue that
> talks about that: https://issues.apache.org/jira/browse/PARQUET-1647
> So, this will not work:
> {code:java}
> import pyarrow as pa
> import pyarrow.parquet as pq
> import numpy as np
> arr = pa.array(np.float16([0.1, 2.2, 3]))
> table = pa.table({'a': arr})
> pq.write_table(table, "test_halffloat.parquet") {code}
> {{This is a proposal to store float16 values in Parquet as FixedSizeBinary,
> and then restore them to float16 when reading them back in.}}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
