[
https://issues.apache.org/jira/browse/LUCENE-10534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17529572#comment-17529572
]
Kevin Risden edited comment on LUCENE-10534 at 4/29/22 5:15 PM:
----------------------------------------------------------------
Updated metrics - there is a benefit to the new maxfloatfunction logic to avoid
duplicate exists() even with using the new fieldsource stuff in LUCENE-10542.
| Benchmark | Mode | Cnt
| Score and Error | Units |
|-----------------------------------------------------------------|-------|-----|------------------|-------|
| MyBenchmark.testMaxFloatFunction | thrpt | 25
| 64.159 ± 2.031 | ops/s |
| MyBenchmark.testNewMaxFloatFunction | thrpt | 25
| 94.997 ± 2.365 | ops/s |
| MyBenchmark.testMaxFloatFunctionNewFloatFieldSource | thrpt | 25
| 123.191 ± 9.291 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionNewFloatFieldSource | thrpt | 25
| 123.817 ± 6.191 | ops/s |
| MyBenchmark.testMaxFloatFunctionRareField | thrpt | 25
| 244.921 ± 6.439 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionRareField | thrpt | 25
| 239.288 ± 5.136 | ops/s |
| MyBenchmark.testMaxFloatFunctionNewFloatFieldSourceRareField | thrpt | 25
| 271.521 ± 3.870 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionNewFloatFieldSourceRareField | thrpt | 25
| 279.334 ± 10.511 | ops/s |
was (Author: risdenk):
Updated metrics - there is a benefit to the new maxfloatfunction logic to avoid
duplicate exists() even with using the new fieldsource stuff in LUCENE-10542.
| Benchmark | Mode | Cnt
| Score and Error | Units |
|-----------------------------------------------------------------|-------|-----|------------------|-------|
| MyBenchmark.testMaxFloatFunction | thrpt | 25
| 69.949 ± 4.043 | ops/s |
| MyBenchmark.testMaxFloatFunctionNewFloatFieldSource | thrpt | 25
| 112.326 ± 3.228 | ops/s |
| MyBenchmark.testNewMaxFloatFunction | thrpt | 25
| 93.216 ± 2.757 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionNewFloatFieldSource | thrpt | 25
| 123.364 ± 7.861 | ops/s |
| MyBenchmark.testMaxFloatFunctionRareField | thrpt | 25
| 257.339 ± 33.849 | ops/s |
| MyBenchmark.testMaxFloatFunctionNewFloatFieldSourceRareField | thrpt | 25
| 287.175 ± 22.840 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionRareField | thrpt | 25
| 235.268 ± 4.103 | ops/s |
| MyBenchmark.testNewMaxFloatFunctionNewFloatFieldSourceRareField | thrpt | 25
| 272.397 ± 8.406 | ops/s |
> MinFloatFunction / MaxFloatFunction exists check can be slow
> ------------------------------------------------------------
>
> Key: LUCENE-10534
> URL: https://issues.apache.org/jira/browse/LUCENE-10534
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Kevin Risden
> Assignee: Kevin Risden
> Priority: Minor
> Attachments: flamegraph.png, flamegraph_getValueForDoc.png
>
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> MinFloatFunction
> (https://github.com/apache/lucene/blob/main/lucene/queries/src/java/org/apache/lucene/queries/function/valuesource/MinFloatFunction.java)
> and MaxFloatFunction
> (https://github.com/apache/lucene/blob/main/lucene/queries/src/java/org/apache/lucene/queries/function/valuesource/MaxFloatFunction.java)
> both check if values exist. This is needed since the underlying valuesource
> returns 0.0f as either a valid value or as a value when the document doesn't
> have a value.
> Even though this is changed to anyExists and short circuits in the case a
> value is found in any document, the worst case is that there is no value
> found and requires checking all the way through to the raw data. This is only
> needed when 0.0f is returned and need to determine if it is a valid value or
> the not found case.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
