[
https://issues.apache.org/jira/browse/PARQUET-41?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16338706#comment-16338706
]
Junjie Chen commented on PARQUET-41:
------------------------------------
In Parquet-mr, when we set dictionary encoding to true, the valueWriter is
FallbackValuesWriter which consist of a dictionaryValueWriter and
FallbackWriter. In the DictionaryValuesWriter, it first determines whether it
should use dictionary encoding by check maxDictionaryBytes which is
DEFAULT_PAGE_SIZE in ParquetProperties.java, and then in FallValuesrWriter, it
also check whether dic index plus dic data is large than raw data, you can see
[line
123|https://github.com/apache/parquet-mr/blob/master/parquet-column/src/main/java/org/apache/parquet/column/values/dictionary/DictionaryValuesWriter.java#L123]
and [line
130|https://github.com/apache/parquet-mr/blob/master/parquet-column/src/main/java/org/apache/parquet/column/values/dictionary/DictionaryValuesWriter.java#L130].
So the column with large cardinality should fallback to plain encoding.
> Add bloom filters to parquet statistics
> ---------------------------------------
>
> Key: PARQUET-41
> URL: https://issues.apache.org/jira/browse/PARQUET-41
> Project: Parquet
> Issue Type: New Feature
> Components: parquet-format, parquet-mr
> Reporter: Alex Levenson
> Assignee: Ferdinand Xu
> Priority: Major
> Labels: filter2
>
> For row groups with no dictionary, we could still produce a bloom filter.
> This could be very useful in filtering entire row groups.
> Pull request:
> https://github.com/apache/parquet-mr/pull/215
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
