Skip to content

Commit

Permalink
Switch TDigestState to use HybridDigest by default (#96904)
Browse files Browse the repository at this point in the history 
* Initial import for TDigest forking.

* Fix MedianTest.

More work needed for TDigestPercentile*Tests and the TDigestTest (and
the rest of the tests) in the tdigest lib to pass.

* Fix Dist.

* Fix AVLTreeDigest.quantile to match Dist for uniform centroids.

* Update docs/changelog/96086.yaml

* Fix `MergingDigest.quantile` to match `Dist` on uniform distribution.

* Add merging to TDigestState.hashCode and .equals.

Remove wrong asserts from tests and MergingDigest.

* Fix style violations for tdigest library.

* Fix typo.

* Fix more style violations.

* Fix more style violations.

* Fix remaining style violations in tdigest library.

* Update results in docs based on the forked tdigest.

* Fix YAML tests in aggs module.

* Fix YAML tests in x-pack/plugin.

* Skip failing V7 compat tests in modules/aggregations.

* Fix TDigest library unittests.

Remove redundant serializing interfaces from the library.

* Remove YAML test versions for older releases.

These tests don't address compatibility issues in mixed cluster tests as
the latter contain a mix of older and newer nodes, so the output depends
on which node is picked as a data node since the forked TDigest library
is not backwards compatible (produces slightly different results).

* Fix test failures in docs and mixed cluster.

* Reduce buffer sizes in MergingDigest to avoid oom.

* Exclude more failing V7 compatibility tests.

* Update results for JdbcCsvSpecIT tests.

* Update results for JdbcDocCsvSpecIT tests.

* Revert unrelated change.

* More test fixes.

* Use version skips instead of blacklisting in mixed cluster tests.

* Switch TDigestState back to AVLTreeDigest.

* Update docs and tests with AVLTreeDigest output.

* Update flaky test.

* Remove dead code, esp around tracking of incoming data.

* Update docs/changelog/96086.yaml

* Delete docs/changelog/96086.yaml

* Remove explicit compression calls.

This was added to prevent concurrency tests from failing, but it leads
to reduces precision. Submit this to see if the concurrency tests are
still failing.

* Revert "Remove explicit compression calls."

This reverts commit 5352c96.

* Remove explicit compression calls to MedianAbsoluteDeviation input.

* Add unittests for AVL and merging digest accuracy.

* Fix spotless violations.

* Delete redundant tests and benchmarks.

* Fix spotless violation.

* Use the old implementation of AVLTreeDigest.

The latest library version is 50% slower and less accurate, as verified
by ComparisonTests.

* Update docs with latest percentile results.

* Update docs with latest percentile results.

* Remove repeated compression calls.

* Update more percentile results.

* Use approximate percentile values in integration tests.

This helps with mixed cluster tests, where some of the tests where
blocked.

* Fix expected percentile value in test.

* Revert in-place node updates in AVL tree.

Update quantile calculations between centroids and min/max values to
match v.3.2.

* Add SortingDigest and HybridDigest.

The SortingDigest tracks all samples in an ArrayList that
gets sorted for quantile calculations. This approach
provides perfectly accurate results and is the most
efficient implementation for up to millions of samples,
at the cost of bloated memory footprint.

The HybridDigest uses a SortingDigest for small sample
populations, then switches to a MergingDigest. This
approach combines to the best performance and results for
small sample counts with very good performance and
acceptable accuracy for effectively unbounded sample
counts.

* Remove deps to the 3.2 library.

* Remove unused licenses for tdigest.

* Revert changes for SortingDigest and HybridDigest.

These will be submitted in a follow-up PR for enabling MergingDigest.

* Remove unused Histogram classes and unit tests.

Delete dead and commented out code, make the remaining tests run
reasonably fast. Remove unused annotations, esp. SuppressWarnings.

* Remove Comparison class, not used.

* Revert "Revert changes for SortingDigest and HybridDigest."

This reverts commit 2336b11.

* Use HybridDigest as default tdigest implementation

Add SortingDigest as a simple structure for percentile calculations that
tracks all data points in a sorted array. This is a fast and perfectly
accurate solution that leads to bloated memory allocation.

Add HybridDigest that uses SortingDigest for small sample counts, then
switches to MergingDigest. This approach delivers extreme
performance and accuracy for small populations while scaling
indefinitely and maintaining acceptable performance and accuracy with
constant memory allocation (15kB by default).

Provide knobs to switch back to AVLTreeDigest, either per query or
through ClusterSettings.

* Small fixes.

* Add javadoc and tests.

* Add javadoc and tests.

* Remove special logic for singletons in the boundaries.

While this helps with the case where the digest contains only
singletons (perfect accuracy), it has a major issue problem
(non-monotonic quantile function) when the first singleton is followed
by a non-singleton centroid. It's preferable to revert to the old
version from 3.2; inaccuracies in a singleton-only digest should be
mitigated by using a sorted array for small sample counts.

* Revert changes to expected values in tests.

This is due to restoring quantile functions to match head.

* Revert changes to expected values in tests.

This is due to restoring quantile functions to match head.

* Tentatively restore percentile rank expected results.

* Use cdf version from 3.2

Update Dist.cdf to use interpolation, use the same cdf
version in AVLTreeDigest and MergingDigest.

* Revert "Tentatively restore percentile rank expected results."

This reverts commit 7718dbb.

* Revert remaining changes compared to main.

* Revert excluded V7 compat tests.

* Exclude V7 compat tests still failing.

* Exclude V7 compat tests still failing.

* Remove ClusterSettings tentatively.

* Initial import for TDigest forking.

* Fix MedianTest.

More work needed for TDigestPercentile*Tests and the TDigestTest (and
the rest of the tests) in the tdigest lib to pass.

* Fix Dist.

* Fix AVLTreeDigest.quantile to match Dist for uniform centroids.

* Update docs/changelog/96086.yaml

* Fix `MergingDigest.quantile` to match `Dist` on uniform distribution.

* Add merging to TDigestState.hashCode and .equals.

Remove wrong asserts from tests and MergingDigest.

* Fix style violations for tdigest library.

* Fix typo.

* Fix more style violations.

* Fix more style violations.

* Fix remaining style violations in tdigest library.

* Update results in docs based on the forked tdigest.

* Fix YAML tests in aggs module.

* Fix YAML tests in x-pack/plugin.

* Skip failing V7 compat tests in modules/aggregations.

* Fix TDigest library unittests.

Remove redundant serializing interfaces from the library.

* Remove YAML test versions for older releases.

These tests don't address compatibility issues in mixed cluster tests as
the latter contain a mix of older and newer nodes, so the output depends
on which node is picked as a data node since the forked TDigest library
is not backwards compatible (produces slightly different results).

* Fix test failures in docs and mixed cluster.

* Reduce buffer sizes in MergingDigest to avoid oom.

* Exclude more failing V7 compatibility tests.

* Update results for JdbcCsvSpecIT tests.

* Update results for JdbcDocCsvSpecIT tests.

* Revert unrelated change.

* More test fixes.

* Use version skips instead of blacklisting in mixed cluster tests.

* Switch TDigestState back to AVLTreeDigest.

* Update docs and tests with AVLTreeDigest output.

* Update flaky test.

* Remove dead code, esp around tracking of incoming data.

* Remove explicit compression calls.

This was added to prevent concurrency tests from failing, but it leads
to reduces precision. Submit this to see if the concurrency tests are
still failing.

* Update docs/changelog/96086.yaml

* Delete docs/changelog/96086.yaml

* Revert "Remove explicit compression calls."

This reverts commit 5352c96.

* Remove explicit compression calls to MedianAbsoluteDeviation input.

* Add unittests for AVL and merging digest accuracy.

* Fix spotless violations.

* Delete redundant tests and benchmarks.

* Fix spotless violation.

* Use the old implementation of AVLTreeDigest.

The latest library version is 50% slower and less accurate, as verified
by ComparisonTests.

* Update docs with latest percentile results.

* Update docs with latest percentile results.

* Remove repeated compression calls.

* Update more percentile results.

* Use approximate percentile values in integration tests.

This helps with mixed cluster tests, where some of the tests where
blocked.

* Fix expected percentile value in test.

* Revert in-place node updates in AVL tree.

Update quantile calculations between centroids and min/max values to
match v.3.2.

* Add SortingDigest and HybridDigest.

The SortingDigest tracks all samples in an ArrayList that
gets sorted for quantile calculations. This approach
provides perfectly accurate results and is the most
efficient implementation for up to millions of samples,
at the cost of bloated memory footprint.

The HybridDigest uses a SortingDigest for small sample
populations, then switches to a MergingDigest. This
approach combines to the best performance and results for
small sample counts with very good performance and
acceptable accuracy for effectively unbounded sample
counts.

* Remove deps to the 3.2 library.

* Remove unused licenses for tdigest.

* Revert changes for SortingDigest and HybridDigest.

These will be submitted in a follow-up PR for enabling MergingDigest.

* Remove unused Histogram classes and unit tests.

Delete dead and commented out code, make the remaining tests run
reasonably fast. Remove unused annotations, esp. SuppressWarnings.

* Remove Comparison class, not used.

* Revert "Revert changes for SortingDigest and HybridDigest."

This reverts commit 2336b11.

* Use HybridDigest as default tdigest implementation

Add SortingDigest as a simple structure for percentile calculations that
tracks all data points in a sorted array. This is a fast and perfectly
accurate solution that leads to bloated memory allocation.

Add HybridDigest that uses SortingDigest for small sample counts, then
switches to MergingDigest. This approach delivers extreme
performance and accuracy for small populations while scaling
indefinitely and maintaining acceptable performance and accuracy with
constant memory allocation (15kB by default).

Provide knobs to switch back to AVLTreeDigest, either per query or
through ClusterSettings.

* Add javadoc and tests.

* Remove ClusterSettings tentatively.

* Restore bySize function in TDigest and subclasses.

* Update Dist.cdf to match the rest.

Update tests.

* Revert outdated test changes.

* Revert outdated changes.

* Small fixes.

* Update docs/changelog/96794.yaml

* TDigestState uses MergingDigest by default.

* Make HybridDigest the default implementation.

* Update boxplot documentation.

* Use HybridDigest for real.

* Restore AVLTreeDigest as the default in TDigestState.

TDigest.createHybridDigest nw returns the right type.
The switch in TDigestState will happen in a separate PR
as it requires many test updates.

* Use execution_hint in tdigest spec.

* Restore expected test values.

* Fix Dist.cdf for empty digest.

* Bump up TransportVersion.

* More test updates.

* Bump up TransportVersion for real.

* Restore V7 compat blacklisting.

* HybridDigest uses its final implementation during deserialization.

* Restore the right TransportVersion in TDigestState.read

* More test fixes.

* More test updates.

* Use TDigestExecutionHint instead of strings.

* Add link to TDigest javadoc.

* Spotless fix.

* Small fixes.

* Bump up TransportVersion.

* Bump up the TransportVersion, again.

* Update docs/changelog/96904.yaml

* Delete 96794.yaml

Delete existing changelog to get a new one.

* Restore previous changelog.

* Rename  96794.yaml to 96794.yaml

* Update breaking change notes in changelog.

* Remove mapping value from changelog.

* Set a valid breaking area.

* Use HybridDigest as default TDigest impl.

* Update docs/changelog/96904.yaml

* Use TDigestExecutionHint in MedianAbsoluteDeviationAggregator.

* Update changelog and comment in blacklisted V7 compat tests.

* Update breaking area in changelog.
  • Loading branch information
@kkrik-es
kkrik-es authored Jun 19, 2023
1 parent ee4fbe4 commit cd3f84c
Show file tree
Hide file tree
Showing 26 changed files with 226 additions and 168 deletions.
18 changes: 18 additions & 0 deletions docs/changelog/96904.yaml
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
pr: 96904
summary: Switch TDigestState to use `HybridDigest` by default
area: Aggregations
type: breaking
issues: []
breaking:
title: Switch TDigestState to use `HybridDigest` by default
area: REST API
details: "The default implementation for TDigest in percentile calculations switches\
\ to a new internal implementation offering superior performance (2x-10x speedup),\
\ at a very small accuracy penalty for very large sample populations."
impact: "This change leads to generating slightly different results in percentile\
\ calculations. If the highest possible accuracy is desired, or it's crucial to\
\ produce exactly the same results as in previous versions, one can either set\
\ `execution_hint` to `high_accuracy` in the `tdigest` spec of a given percentile\
\ calculation, or set `search.aggs.tdigest_execution_hint` to `high_accuracy`\
\ in cluster settings to apply to all percentile queries."
notable: true
8 changes: 4 additions & 4 deletions docs/reference/aggregations/metrics/boxplot-aggregation.asciidoc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -55,9 +55,9 @@ The response will look like this:
"load_time_boxplot": {
"min": 0.0,
"max": 990.0,
"q1": 165.0,
"q1": 167.5,
"q2": 445.0,
"q3": 725.0,
"q3": 722.5,
"lower": 0.0,
"upper": 990.0
}
Expand Down Expand Up @@ -114,9 +114,9 @@ GET latency/_search
"load_time_boxplot": {
"min": 0.0,
"max": 99.0,
"q1": 16.5,
"q1": 16.75,
"q2": 44.5,
"q3": 72.5,
"q3": 72.25,
"lower": 0.0,
"upper": 99.0
}
Expand Down
50 changes: 32 additions & 18 deletions docs/reference/aggregations/metrics/percentile-aggregation.asciidoc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -54,19 +54,26 @@ percentiles: `[ 1, 5, 25, 50, 75, 95, 99 ]`. The response will look like this:
"aggregations": {
"load_time_outlier": {
"values": {
"1.0": 5.0,
"5.0": 25.0,
"25.0": 165.0,
"1.0": 10.0,
"5.0": 30.0,
"25.0": 170.0,
"50.0": 445.0,
"75.0": 725.0,
"95.0": 945.0,
"99.0": 985.0
"75.0": 720.0,
"95.0": 940.0,
"99.0": 980.0
}
}
}
}
--------------------------------------------------
// TESTRESPONSE[s/\.\.\./"took": $body.took,"timed_out": false,"_shards": $body._shards,"hits": $body.hits,/]
// TESTRESPONSE[s/"1.0": 10.0/"1.0": 9.9/]
// TESTRESPONSE[s/"5.0": 30.0/"5.0": 29.5/]
// TESTRESPONSE[s/"25.0": 170.0/"25.0": 167.5/]
// TESTRESPONSE[s/"50.0": 445.0/"50.0": 445.0/]
// TESTRESPONSE[s/"75.0": 720.0/"75.0": 722.5/]
// TESTRESPONSE[s/"95.0": 940.0/"95.0": 940.5/]
// TESTRESPONSE[s/"99.0": 980.0/"99.0": 980.1/]

As you can see, the aggregation will return a calculated value for each percentile
in the default range. If we assume response times are in milliseconds, it is
Expand Down Expand Up @@ -128,38 +135,45 @@ Response:
"values": [
{
"key": 1.0,
"value": 5.0
"value": 10.0
},
{
"key": 5.0,
"value": 25.0
"value": 30.0
},
{
"key": 25.0,
"value": 165.0
"value": 170.0
},
{
"key": 50.0,
"value": 445.0
},
{
"key": 75.0,
"value": 725.0
"value": 720.0
},
{
"key": 95.0,
"value": 945.0
"value": 940.0
},
{
"key": 99.0,
"value": 985.0
"value": 980.0
}
]
}
}
}
--------------------------------------------------
// TESTRESPONSE[s/\.\.\./"took": $body.took,"timed_out": false,"_shards": $body._shards,"hits": $body.hits,/]
// TESTRESPONSE[s/"value": 10.0/"value": 9.9/]
// TESTRESPONSE[s/"value": 30.0/"value": 29.5/]
// TESTRESPONSE[s/"value": 170.0/"value": 167.5/]
// TESTRESPONSE[s/"value": 445.0/"value": 445.0/]
// TESTRESPONSE[s/"value": 720.0/"value": 722.5/]
// TESTRESPONSE[s/"value": 940.0/"value": 940.5/]
// TESTRESPONSE[s/"value": 980.0/"value": 980.1/]

==== Script

Expand Down Expand Up @@ -203,13 +217,13 @@ GET latency/_search
"aggregations": {
"load_time_outlier": {
"values": {
"1.0": 0.5,
"5.0": 2.5,
"25.0": 16.5,
"1.0": 0.99,
"5.0": 2.95,
"25.0": 16.75,
"50.0": 44.5,
"75.0": 72.5,
"95.0": 94.5,
"99.0": 98.5
"75.0": 72.25,
"95.0": 94.05,
"99.0": 98.01
}
}
}
Expand Down
14 changes: 7 additions & 7 deletions docs/reference/aggregations/pipeline/moving-percentiles-aggregation.asciidoc
View file
Original file line number Diff line number Diff line change
Expand Up @@ -97,7 +97,7 @@ And the following may be the response:
"doc_count": 3,
"the_percentile": {
"values": {
"1.0": 150.0,
"1.0": 151.0,
"99.0": 200.0
}
}
Expand All @@ -108,13 +108,13 @@ And the following may be the response:
"doc_count": 2,
"the_percentile": {
"values": {
"1.0": 10.0,
"99.0": 50.0
"1.0": 10.4,
"99.0": 49.6
}
},
"the_movperc": {
"values": {
"1.0": 150.0,
"1.0": 151.0,
"99.0": 200.0
}
}
Expand All @@ -125,13 +125,13 @@ And the following may be the response:
"doc_count": 2,
"the_percentile": {
"values": {
"1.0": 175.0,
"99.0": 200.0
"1.0": 175.25,
"99.0": 199.75
}
},
"the_movperc": {
"values": {
"1.0": 10.0,
"1.0": 11.6,
"99.0": 200.0
}
}
Expand Down
12 changes: 11 additions & 1 deletion modules/aggregations/build.gradle
View file
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,17 @@ tasks.named("yamlRestTestV7CompatTransform").configure { task ->
task.skipTest("search.aggregation/20_terms/numeric profiler", "The profiler results aren't backwards compatible.")
task.skipTest("search.aggregation/210_top_hits_nested_metric/top_hits aggregation with sequence numbers", "#42809 the use nested path and filter sort throws an exception")
task.skipTest("search.aggregation/370_doc_count_field/Test filters agg with doc_count", "Uses profiler for assertions which is not backwards compatible")
task.skipTest("search.aggregation/420_percentile_ranks_tdigest_metric/filtered", "Uses t-digest library which is not backwards compatible")

// In 8.9.0, the default t-digest algorithm changed from AVL-tree-based to hybrid, combining a sorted array of samples with a merging
// implementation. This change leads to slight different percentile results, compared to previous versions.
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Basic test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Non-keyed test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Only aggs test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Explicit Percents test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Metadata test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/180_percentiles_tdigest_metric/Filtered test", "Hybrid t-digest produces different results.")
task.skipTest("search.aggregation/420_percentile_ranks_tdigest_metric/filtered", "Hybrid t-digest produces different results.")

task.addAllowedWarningRegex("\\[types removal\\].*")
}

Expand Down
51 changes: 27 additions & 24 deletions ...src/yamlRestTest/resources/rest-api-spec/test/aggregations/percentiles_tdigest_metric.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -46,6 +46,9 @@ setup:

---
"Basic test":
- skip:
version: " - 8.8.99"
reason: t-digest forked in 8.9.0
- do:
search:
rest_total_hits_as_int: true
Expand All @@ -61,21 +64,21 @@ setup:
- match: { hits.total: 4 }
- length: { hits.hits: 4 }

- match: { aggregations.percentiles_int.values.1\.0: 1.0 }
- match: { aggregations.percentiles_int.values.5\.0: 1.0 }
- match: { aggregations.percentiles_int.values.25\.0: 26.0 }
- match: { aggregations.percentiles_int.values.1\.0: 2.5 }
- match: { aggregations.percentiles_int.values.5\.0: 8.5 }
- match: { aggregations.percentiles_int.values.25\.0: 38.5 }
- match: { aggregations.percentiles_int.values.50\.0: 76.0 }
- match: { aggregations.percentiles_int.values.75\.0: 126.0 }
- match: { aggregations.percentiles_int.values.95\.0: 151.0 }
- match: { aggregations.percentiles_int.values.99\.0: 151.0 }
- match: { aggregations.percentiles_int.values.75\.0: 113.5 }
- match: { aggregations.percentiles_int.values.95\.0: 143.5 }
- match: { aggregations.percentiles_int.values.99\.0: 149.5 }

- match: { aggregations.percentiles_double.values.1\.0: 1.0 }
- match: { aggregations.percentiles_double.values.5\.0: 1.0 }
- match: { aggregations.percentiles_double.values.25\.0: 26.0 }
- match: { aggregations.percentiles_double.values.1\.0: 2.5 }
- match: { aggregations.percentiles_double.values.5\.0: 8.5 }
- match: { aggregations.percentiles_double.values.25\.0: 38.5 }
- match: { aggregations.percentiles_double.values.50\.0: 76.0 }
- match: { aggregations.percentiles_double.values.75\.0: 126.0 }
- match: { aggregations.percentiles_double.values.95\.0: 151.0 }
- match: { aggregations.percentiles_double.values.99\.0: 151.0 }
- match: { aggregations.percentiles_double.values.75\.0: 113.5 }
- match: { aggregations.percentiles_double.values.95\.0: 143.5 }
- match: { aggregations.percentiles_double.values.99\.0: 149.5 }

- do:
search:
Expand All @@ -97,21 +100,21 @@ setup:
- match: { hits.total: 4 }
- length: { hits.hits: 4 }

- match: { aggregations.percentiles_int.values.1\.0: 1.0 }
- match: { aggregations.percentiles_int.values.5\.0: 1.0 }
- match: { aggregations.percentiles_int.values.25\.0: 26.0 }
- match: { aggregations.percentiles_int.values.1\.0: 2.5 }
- match: { aggregations.percentiles_int.values.5\.0: 8.5 }
- match: { aggregations.percentiles_int.values.25\.0: 38.5 }
- match: { aggregations.percentiles_int.values.50\.0: 76.0 }
- match: { aggregations.percentiles_int.values.75\.0: 126.0 }
- match: { aggregations.percentiles_int.values.95\.0: 151.0 }
- match: { aggregations.percentiles_int.values.99\.0: 151.0 }
- match: { aggregations.percentiles_int.values.75\.0: 113.5 }
- match: { aggregations.percentiles_int.values.95\.0: 143.5 }
- match: { aggregations.percentiles_int.values.99\.0: 149.5 }

- match: { aggregations.percentiles_double.values.1\.0: 1.0 }
- match: { aggregations.percentiles_double.values.5\.0: 1.0 }
- match: { aggregations.percentiles_double.values.25\.0: 26.0 }
- match: { aggregations.percentiles_double.values.1\.0: 2.5 }
- match: { aggregations.percentiles_double.values.5\.0: 8.5 }
- match: { aggregations.percentiles_double.values.25\.0: 38.5 }
- match: { aggregations.percentiles_double.values.50\.0: 76.0 }
- match: { aggregations.percentiles_double.values.75\.0: 126.0 }
- match: { aggregations.percentiles_double.values.95\.0: 151.0 }
- match: { aggregations.percentiles_double.values.99\.0: 151.0 }
- match: { aggregations.percentiles_double.values.75\.0: 113.5 }
- match: { aggregations.percentiles_double.values.95\.0: 143.5 }
- match: { aggregations.percentiles_double.values.99\.0: 149.5 }


---
Expand Down
2 changes: 1 addition & 1 deletion ...java/org/elasticsearch/search/aggregations/metrics/MedianAbsoluteDeviationAggregator.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 +92,7 @@ public void collect(int doc, long bucket) throws IOException {

TDigestState valueSketch = valueSketches.get(bucket);
if (valueSketch == null) {
valueSketch = TDigestState.create(compression);
valueSketch = TDigestState.create(compression, executionHint);
valueSketches.set(bucket, valueSketch);
}

Expand Down
2 changes: 1 addition & 1 deletion server/src/main/java/org/elasticsearch/search/aggregations/metrics/TDigestState.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ protected enum Type {
SORTING;

static Type defaultValue() {
return AVL_TREE;
return HYBRID;
}

static Type valueForHighAccuracy() {
Expand Down
14 changes: 7 additions & 7 deletions ...java/org/elasticsearch/search/aggregations/metrics/TDigestPercentilesAggregatorTests.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -83,8 +83,8 @@ public void testSomeMatchesSortedNumericDocValues() throws IOException {
}, tdigest -> {
assertEquals(7L, tdigest.state.size());
assertEquals(7L, tdigest.state.centroidCount());
assertEquals(4.5d, tdigest.percentile(75), 0.0d);
assertEquals("4.5", tdigest.percentileAsString(75));
assertEquals(4.0d, tdigest.percentile(75), 0.0d);
assertEquals("4.0", tdigest.percentileAsString(75));
assertEquals(2.0d, tdigest.percentile(50), 0.0d);
assertEquals("2.0", tdigest.percentileAsString(50));
assertEquals(1.0d, tdigest.percentile(22), 0.0d);
Expand All @@ -107,14 +107,14 @@ public void testSomeMatchesNumericDocValues() throws IOException {
assertEquals(tdigest.state.centroidCount(), 7L);
assertEquals(8.0d, tdigest.percentile(100), 0.0d);
assertEquals("8.0", tdigest.percentileAsString(100));
assertEquals(6.98d, tdigest.percentile(88), 0.0d);
assertEquals("6.98", tdigest.percentileAsString(88));
assertEquals(4.0d, tdigest.percentile(75), 0.0d);
assertEquals("4.0", tdigest.percentileAsString(75));
assertEquals(1.0d, tdigest.percentile(33), 0.0d);
assertEquals("1.0", tdigest.percentileAsString(33));
assertEquals(1.0d, tdigest.percentile(25), 0.0d);
assertEquals("1.0", tdigest.percentileAsString(25));
assertEquals(0.0d, tdigest.percentile(1), 0.0d);
assertEquals("0.0", tdigest.percentileAsString(1));
assertEquals(0.06d, tdigest.percentile(1), 0.0d);
assertEquals("0.06", tdigest.percentileAsString(1));
assertTrue(AggregationInspectionHelper.hasValue(tdigest));
});
}
Expand All @@ -135,7 +135,7 @@ public void testQueryFiltering() throws IOException {
assertEquals(4L, tdigest.state.centroidCount());
assertEquals(2.0d, tdigest.percentile(100), 0.0d);
assertEquals(1.0d, tdigest.percentile(50), 0.0d);
assertEquals(0.5d, tdigest.percentile(25), 0.0d);
assertEquals(0.75d, tdigest.percentile(25), 0.0d);
assertTrue(AggregationInspectionHelper.hasValue(tdigest));
});

Expand Down
3 changes: 3 additions & 0 deletions server/src/test/java/org/elasticsearch/search/aggregations/metrics/TDigestStateTests.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -158,6 +158,8 @@ public void testFactoryMethods() {

assertEquals(fast, anotherFast);
assertEquals(accurate, anotherAccurate);
assertNotEquals(fast, accurate);
assertNotEquals(anotherFast, anotherAccurate);
}

private static TDigestState writeToAndReadFrom(TDigestState state, TransportVersion version) throws IOException {
Expand Down Expand Up @@ -195,6 +197,7 @@ public void testSerialization() throws IOException {
assertEquals(serialized, state);

TDigestState serializedBackwardsCompatible = writeToAndReadFrom(state, TransportVersion.V_8_500_000);
assertNotEquals(serializedBackwardsCompatible, state);
assertEquals(serializedBackwardsCompatible, backwardsCompatible);
}
}
10 changes: 5 additions & 5 deletions .../xpack/analytics/aggregations/metrics/TDigestPreAggregatedPercentilesAggregatorTests.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -88,10 +88,10 @@ public void testSomeMatchesBinaryDocValues() throws IOException {
}, hdr -> {
// assertEquals(4L, hdr.state.getTotalCount());
double approximation = 0.05d;
assertEquals(15.0d, hdr.percentile(25), approximation);
assertEquals(17.5d, hdr.percentile(25), approximation);
assertEquals(30.0d, hdr.percentile(50), approximation);
assertEquals(50.0d, hdr.percentile(75), approximation);
assertEquals(60.0d, hdr.percentile(99), approximation);
assertEquals(45.0d, hdr.percentile(75), approximation);
assertEquals(59.4d, hdr.percentile(99), approximation);
assertTrue(AggregationInspectionHelper.hasValue(hdr));
});
}
Expand All @@ -105,9 +105,9 @@ public void testSomeMatchesMultiBinaryDocValues() throws IOException {
}, hdr -> {
// assertEquals(16L, hdr.state.getTotalCount());
double approximation = 0.05d;
assertEquals(15.0d, hdr.percentile(25), approximation);
assertEquals(17.5d, hdr.percentile(25), approximation);
assertEquals(30.0d, hdr.percentile(50), approximation);
assertEquals(50.0d, hdr.percentile(75), approximation);
assertEquals(45.0d, hdr.percentile(75), approximation);
assertEquals(60.0d, hdr.percentile(99), approximation);
assertTrue(AggregationInspectionHelper.hasValue(hdr));
});
Expand Down
Loading

0 comments on commit cd3f84c

Please sign in to comment.