[GitHub] [lucene] mayya-sharipova commented on a diff in pull request #11743: LUCENE-10592 Better estimate memory for HNSW graph

mayya-sharipova commented on code in PR #11743: URL: https://github.com/apache/lucene/pull/11743#discussion_r962072588 ## lucene/core/src/java/org/apache/lucene/util/hnsw/OnHeapHnswGraph.java: ## @@ -175,20 +175,28 @@ public long ramBytesUsed() { long neighborArrayBytes0 =

[GitHub] [lucene] gsmiller commented on issue #11553: Improve performance of SortedSetDV faceting by iterating on class types [LUCENE-10517]

gsmiller commented on issue #11553: URL: https://github.com/apache/lucene/issues/11553#issuecomment-1235993045 @ChrisHegarty I was looking back through faceting improvements and trying to catch up on this change, which I hadn't followed closely when it was being made. When you get a chance

[GitHub] [lucene] mayya-sharipova commented on a diff in pull request #11743: LUCENE-10592 Better estimate memory for HNSW graph

mayya-sharipova commented on code in PR #11743: URL: https://github.com/apache/lucene/pull/11743#discussion_r962071131 ## lucene/core/src/java/org/apache/lucene/util/hnsw/NeighborArray.java: ## @@ -104,8 +104,8 @@ public void removeLast() { } public void removeIndex(int

[GitHub] [lucene] mayya-sharipova commented on a diff in pull request #11743: LUCENE-10592 Better estimate memory for HNSW graph

mayya-sharipova commented on code in PR #11743: URL: https://github.com/apache/lucene/pull/11743#discussion_r962070814 ## lucene/core/src/test/org/apache/lucene/util/TestRamUsageEstimator.java: ## @@ -222,6 +229,33 @@ public void testPrintValues() { System.out.println("LONG

[GitHub] [lucene] msokolov opened a new pull request, #11747: update DOAP and releaseWizard to reflect migration to github

msokolov opened a new pull request, #11747: URL: https://github.com/apache/lucene/pull/11747 While preparing the 9.4.0 release, I ran across some references to JIRA and update those. I also found that the region names used in the holiday.py module installed by pip for me were different than

[GitHub] [lucene] gsmiller merged pull request #11737: Simplify dense optimization check in TermInSetQuery

gsmiller merged PR #11737: URL: https://github.com/apache/lucene/pull/11737 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.ap

[GitHub] [lucene] msokolov commented on a diff in pull request #11743: LUCENE-10592 Better estimate memory for HNSW graph

msokolov commented on code in PR #11743: URL: https://github.com/apache/lucene/pull/11743#discussion_r961747487 ## lucene/core/src/java/org/apache/lucene/util/hnsw/NeighborArray.java: ## @@ -104,8 +104,8 @@ public void removeLast() { } public void removeIndex(int idx) {

[GitHub] [lucene] gsmiller opened a new pull request, #11746: Deprecate LongValueFacetCounts#getTopChildrenSortByCount since it provides redundant functionality

gsmiller opened a new pull request, #11746: URL: https://github.com/apache/lucene/pull/11746 ### Description This is a companion (backport) PR to #11744 that marks functionality deprecated instead of outright removing it. -- This is an automated message from the Apache Git Serv

[GitHub] [lucene] rishabhmaurya opened a new issue, #11745: Store summarized results in internal nodes of BKD for time series points

rishabhmaurya opened a new issue, #11745: URL: https://github.com/apache/lucene/issues/11745 ### Description Time series points have a timestamp, measurement and dimensions associated with them. The common queries are range queries on timestamp, metric aggregation on measurement and

[GitHub] [lucene] gsmiller opened a new pull request, #11744: Remove LongValueFacetCounts#getTopChildrenSortByCount since it provides redundant functionality

gsmiller opened a new pull request, #11744: URL: https://github.com/apache/lucene/pull/11744 ### Description `LongValueFacetCounts#getTopChildrenSortByCount` does exactly the same thing as the more standard `LongValueFacetCounts#getTopChildren`, so we can tighten up our API. N

[GitHub] [lucene] mayya-sharipova opened a new pull request, #11743: LUCENE-10592 Better estimate memory for HNSW graph

mayya-sharipova opened a new pull request, #11743: URL: https://github.com/apache/lucene/pull/11743 Better estimate memory used for OnHeapHnswGraph, as well as add tests. Also don't over-allocate arrays in NeighborArray. Relates to #992 -- This is an automated message from

[GitHub] [lucene] gsmiller closed issue #11574: TopN is not being used in getTopChildren() [LUCENE-10538]

gsmiller closed issue #11574: TopN is not being used in getTopChildren() [LUCENE-10538] URL: https://github.com/apache/lucene/issues/11574 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specifi

[GitHub] [lucene] gsmiller commented on issue #11574: TopN is not being used in getTopChildren() [LUCENE-10538]

gsmiller commented on issue #11574: URL: https://github.com/apache/lucene/issues/11574#issuecomment-1235534712 Let's resolve this out. This issue grew into a few different spin-offs, which are now all resolved. But... it looks like we introduced another case of not properly implementing "to

[GitHub] [lucene] gsmiller opened a new issue, #11742: MatchingFacetSetsCounts doesn't properly implement getTopChildren

gsmiller opened a new issue, #11742: URL: https://github.com/apache/lucene/issues/11742 ### Description `MatchingFacetSetsCounts#getTopChildren` is currently just delegating to `#getAllChildren`, which isn't really the correct thing to do. We should properly implement "top children."