Skip to content

Conversation

qurbonzoda
Copy link
Contributor

Benchmark parameters adjustment and benchmark result files will be removed before merging.

Introducing CollisionTrieNode for collision nodes (polymorphism approach) adds significant performance losses, see hashSet regressions: https://teamcity.jetbrains.com/viewLog.html?buildId=2469110&buildTypeId=KotlinTools_KotlinxCollectionsImmutable_BenchmarkHashIterations&tab=artifacts#%2FbenchmarkResults
Thus, collision nodes and normal nodes were not separated into different entities.

Final benchmark results: https://teamcity.jetbrains.com/viewLog.html?buildId=2469130&buildTypeId=KotlinTools_KotlinxCollectionsImmutable_BenchmarkHashIterations&tab=artifacts

@qurbonzoda qurbonzoda requested a review from ilya-g August 8, 2019 19:25
@qurbonzoda qurbonzoda force-pushed the set-and-map-collisions-optimization branch from 89f6700 to 6b0db5e Compare August 9, 2019 18:11
@qurbonzoda
Copy link
Contributor Author

Also addresses #45.

@ilya-g
Copy link
Member

ilya-g commented Oct 1, 2019

Let's try to eliminate asymptotic overhead on big sets/maps. For example, given the current representation of a collision node, I believe isCollision could be checked only in leaf nodes on the operation path.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants