[
https://issues.apache.org/jira/browse/ARROW-10173?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17230223#comment-17230223
]
Yordan Pavlov commented on ARROW-10173:
---------------------------------------
I have an initial implementation of direct comparison operations to scalar
values in datafusion which shows promising results.
For a simple query like this: "select f32, f64 from t where f32 >= 250 and f64
> 250"
the old, all array implementation takes this long:
filter_20_12 time: [74.186 ms 76.356 ms 78.572 ms]
where as the new implementation shows approximately 10x performance improvement:
filter_20_12 time: [6.5256 ms 6.5701 ms 6.6171 ms]
change: [-91.173% -90.996% -90.820%] (p = 0.00 < 0.05)
I will be submitting a PR for this change soon.
> [Rust][DataFusion] Improve performance of equality to a constant predicate
> support
> ----------------------------------------------------------------------------------
>
> Key: ARROW-10173
> URL: https://issues.apache.org/jira/browse/ARROW-10173
> Project: Apache Arrow
> Issue Type: Improvement
> Reporter: Andrew Lamb
> Priority: Major
>
> I noticed this behavior while working on support for DictionaryArrays and
> wanted to capture it in a ticket in case someone has time to work on it.
> In order to implement an equality predicate to a constant such as {{d1 =
> 'three'}}, DataFusion effectively creates an array with the same value
> {{'three'}} repeated over and over again and uses the equality compute
> kernel. This is ... suboptimal.
> Here is what the predicate looks like:
> {code}
> predicate: BinaryExpr {
> left: CastExpr {
> expr: Column {
> name: "d1",
> },
> cast_type: Utf8,
> },
> op: Eq,
> right: Literal {
> value: Utf8("three"),
> },
> },
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
