[ML] adding new KS test pipeline aggregation #73334
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change | ||||
|---|---|---|---|---|---|---|
| @@ -0,0 +1,292 @@ | ||||||
| [role="xpack"] | ||||||
| [testenv="basic"] | ||||||
| [[search-aggregations-bucket-count-ks-test-aggregation]] | ||||||
| === Bucket count K-S test correlation aggregation | ||||||
| ++++ | ||||||
| <titleabbrev>Bucket count K-S test aggregation</titleabbrev> | ||||||
| ++++ | ||||||
|
|
||||||
| experimental::[] | ||||||
|
|
||||||
| A sibling pipeline aggregation which executes a two sample Kolmogorov–Smirnov test | ||||||
| (referred to as a "K-S test" from now on) against a provided distribution and | ||||||
| the distribution of documents counts in the configured sibling aggregation. | ||||||
|
|
||||||
| This test is useful to determine if two samples (represented by `fractions` and `buckets_path`) are | ||||||
| drawn from the same distribution. | ||||||
|
|
||||||
| [[bucket-count-ks-test-agg-syntax]] | ||||||
| ==== Parameters | ||||||
|
|
||||||
| `buckets_path`:: | ||||||
| (Required, string) | ||||||
| Path to the buckets that contain one set of values to correlate. Must be a `_count` path | ||||||
| For syntax, see <<buckets-path-syntax>>. | ||||||
|
|
||||||
| `alternative`:: | ||||||
| (Optional, list) | ||||||
| A list of string values indicating which K-S test alternative to calculate. | ||||||
| The valid values are: "greater", "less", "two_sided". This parameter is key for | ||||||
| determining the K-S statistic used when calculating the K-S test. Default value is | ||||||
| all possible alternative hypothesis. | ||||||
|
There was a problem hiding this comment. typo
Suggested change
|
||||||
|
|
||||||
| `fractions`:: | ||||||
| (Optional, list) | ||||||
| A list of doubles indicating the distribution of the samples with which to compare to the | ||||||
| `buckets_path` results. The default is a uniform distribution of the same length as the | ||||||
| `buckets_path` buckets. | ||||||
|
There was a problem hiding this comment. We have rather specific requirements on what fractions mean. To produce a meaningful result from this aggregation they should be related to some metric distribution which is then used to create the sibling aggregation. A natural choice is to use equal percentile range queries to construct the sibling aggregation in which case the default is correct. I think it is worth capturing something along these lines. There was a problem hiding this comment. I'd propose something like: |
||||||
|
|
||||||
| `sampling_method`:: | ||||||
| (Optional, string) | ||||||
| Indicates the sampling methodology when calculating the K-S test. Note, this is sampling | ||||||
| of the returned values. This determines the cumulative distribution function (cdf) points | ||||||
|
There was a problem hiding this comment.
Suggested change
Capitalised |
||||||
| used comparing the two samples. Default is `upper_tail`, which emphasizes the upper | ||||||
| end of the CDF points. Valid options are: `upper_tail`, `uniform`, and `lower_tail`. | ||||||
|
|
||||||
| ==== Syntax | ||||||
|
|
||||||
| A `bucket_count_ks_test` aggregation looks like this in isolation: | ||||||
|
|
||||||
| [source,js] | ||||||
| -------------------------------------------------- | ||||||
| { | ||||||
| "bucket_count_ks_test": { | ||||||
| "buckets_path": "range_values>_count", <1> | ||||||
| "alternative": ["less", "greater", "two_sided"], <2> | ||||||
| "sampling_method": "upper_tail" <3> | ||||||
| } | ||||||
| } | ||||||
| -------------------------------------------------- | ||||||
| // NOTCONSOLE | ||||||
| <1> The buckets containing the values to test against. | ||||||
davidkyle marked this conversation as resolved.
Show resolved
Hide resolved
|
||||||
| <2> The alternatives to calculate. | ||||||
| <3> The sampling method for the K-S statistic. | ||||||
|
|
||||||
|
|
||||||
| [[bucket-count-ks-test-agg-example]] | ||||||
| ==== Example | ||||||
|
|
||||||
| The following snippet runs the `bucket_count_ks_test` on the individual terms in the field `version` against a uniform distribution. | ||||||
| The uniform distribution reflects the `latency` percentile buckets. Not shown is the pre-calculation of the `latency` indicator values, | ||||||
| which was done utilizing the | ||||||
| <<search-aggregations-metrics-percentile-aggregation,percentiles>> aggregation. | ||||||
|
|
||||||
| This example is only using the 10s percentiles. | ||||||
|
There was a problem hiding this comment.
Suggested change
|
||||||
|
|
||||||
| [source,console] | ||||||
| ------------------------------------------------- | ||||||
| POST correlate_latency/_search?size=0&filter_path=aggregations | ||||||
| { | ||||||
| "aggs": { | ||||||
| "buckets": { | ||||||
| "terms": { <1> | ||||||
| "field": "version", | ||||||
| "size": 2 | ||||||
| }, | ||||||
| "aggs": { | ||||||
| "latency_ranges": { | ||||||
| "range": { <2> | ||||||
| "field": "latency", | ||||||
| "ranges": [ | ||||||
| { "to": 0 }, | ||||||
| { "from": 0, "to": 105 }, | ||||||
| { "from": 105, "to": 225 }, | ||||||
| { "from": 225, "to": 445 }, | ||||||
| { "from": 445, "to": 665 }, | ||||||
| { "from": 665, "to": 885 }, | ||||||
| { "from": 885, "to": 1115 }, | ||||||
| { "from": 1115, "to": 1335 }, | ||||||
| { "from": 1335, "to": 1555 }, | ||||||
| { "from": 1555, "to": 1775 }, | ||||||
| { "from": 1775 } | ||||||
| ] | ||||||
| } | ||||||
| }, | ||||||
| "ks_test": { <3> | ||||||
| "bucket_count_ks_test": { | ||||||
| "buckets_path": "latency_ranges>_count", | ||||||
| "alternative": ["less", "greater", "two_sided"] | ||||||
| } | ||||||
| } | ||||||
| } | ||||||
| } | ||||||
| } | ||||||
| } | ||||||
| ------------------------------------------------- | ||||||
| // TEST[setup:correlate_latency] | ||||||
|
|
||||||
| <1> The term buckets containing a range aggregation and the bucket correlation aggregation. Both are utilized to calculate | ||||||
| the correlation of the term values with the latency. | ||||||
| <2> The range aggregation on the latency field. The ranges were created referencing the percentiles of the latency field. | ||||||
|
There was a problem hiding this comment. It would be nice to have a full example somewhere in docs of showing how the ranges were found There was a problem hiding this comment. Possibly, I think though that is a larger "how correlations + K-S tests work" type of docs. There was a problem hiding this comment. Actually this isn't so bad as you explain that the latency ranges were calculated using a percentiles agg |
||||||
| <3> The bucket count K-S test aggregation that tests if the bucket counts comes from the same distribution as `fractions`; | ||||||
| where `fractions` is a uniform distribution. | ||||||
|
|
||||||
| And the following may be the response: | ||||||
|
|
||||||
| [source,console-result] | ||||||
| ---- | ||||||
| { | ||||||
| "aggregations" : { | ||||||
| "buckets" : { | ||||||
| "doc_count_error_upper_bound" : 0, | ||||||
| "sum_other_doc_count" : 0, | ||||||
| "buckets" : [ | ||||||
| { | ||||||
| "key" : "1.0", | ||||||
| "doc_count" : 100, | ||||||
| "latency_ranges" : { | ||||||
| "buckets" : [ | ||||||
| { | ||||||
| "key" : "*-0.0", | ||||||
| "to" : 0.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "0.0-105.0", | ||||||
| "from" : 0.0, | ||||||
| "to" : 105.0, | ||||||
| "doc_count" : 1 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "105.0-225.0", | ||||||
| "from" : 105.0, | ||||||
| "to" : 225.0, | ||||||
| "doc_count" : 9 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "225.0-445.0", | ||||||
| "from" : 225.0, | ||||||
| "to" : 445.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "445.0-665.0", | ||||||
| "from" : 445.0, | ||||||
| "to" : 665.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "665.0-885.0", | ||||||
| "from" : 665.0, | ||||||
| "to" : 885.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "885.0-1115.0", | ||||||
| "from" : 885.0, | ||||||
| "to" : 1115.0, | ||||||
| "doc_count" : 10 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1115.0-1335.0", | ||||||
| "from" : 1115.0, | ||||||
| "to" : 1335.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1335.0-1555.0", | ||||||
| "from" : 1335.0, | ||||||
| "to" : 1555.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1555.0-1775.0", | ||||||
| "from" : 1555.0, | ||||||
| "to" : 1775.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1775.0-*", | ||||||
| "from" : 1775.0, | ||||||
| "doc_count" : 20 | ||||||
| } | ||||||
| ] | ||||||
| }, | ||||||
| "ks_test" : { | ||||||
| "less" : 2.248673241788478E-4, | ||||||
| "greater" : 1.0, | ||||||
| "two_sided" : 2.248673241788478E-4 | ||||||
| } | ||||||
| }, | ||||||
| { | ||||||
| "key" : "2.0", | ||||||
| "doc_count" : 100, | ||||||
| "latency_ranges" : { | ||||||
| "buckets" : [ | ||||||
| { | ||||||
| "key" : "*-0.0", | ||||||
| "to" : 0.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "0.0-105.0", | ||||||
| "from" : 0.0, | ||||||
| "to" : 105.0, | ||||||
| "doc_count" : 19 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "105.0-225.0", | ||||||
| "from" : 105.0, | ||||||
| "to" : 225.0, | ||||||
| "doc_count" : 11 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "225.0-445.0", | ||||||
| "from" : 225.0, | ||||||
| "to" : 445.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "445.0-665.0", | ||||||
| "from" : 445.0, | ||||||
| "to" : 665.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "665.0-885.0", | ||||||
| "from" : 665.0, | ||||||
| "to" : 885.0, | ||||||
| "doc_count" : 20 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "885.0-1115.0", | ||||||
| "from" : 885.0, | ||||||
| "to" : 1115.0, | ||||||
| "doc_count" : 10 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1115.0-1335.0", | ||||||
| "from" : 1115.0, | ||||||
| "to" : 1335.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1335.0-1555.0", | ||||||
| "from" : 1335.0, | ||||||
| "to" : 1555.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1555.0-1775.0", | ||||||
| "from" : 1555.0, | ||||||
| "to" : 1775.0, | ||||||
| "doc_count" : 0 | ||||||
| }, | ||||||
| { | ||||||
| "key" : "1775.0-*", | ||||||
| "from" : 1775.0, | ||||||
| "doc_count" : 0 | ||||||
| } | ||||||
| ] | ||||||
| }, | ||||||
| "ks_test" : { | ||||||
| "less" : 0.9642895789647244, | ||||||
| "greater" : 4.58718174664754E-9, | ||||||
| "two_sided" : 4.58718174664754E-9 | ||||||
|
There was a problem hiding this comment. I would expect There was a problem hiding this comment. Yeah, the |
||||||
| } | ||||||
| } | ||||||
| ] | ||||||
| } | ||||||
| } | ||||||
| } | ||||||
| ---- | ||||||
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,43 @@ | ||
| /* | ||
| * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one | ||
| * or more contributor license agreements. Licensed under the Elastic License | ||
| * 2.0; you may not use this file except in compliance with the Elastic License | ||
| * 2.0. | ||
| */ | ||
|
|
||
| package org.elasticsearch.xpack.ml.aggs; | ||
|
|
||
| import org.elasticsearch.common.Numbers; | ||
|
|
||
| public final class DoubleArray { | ||
|
There was a problem hiding this comment. Could you add a unit test for this class? |
||
|
|
||
| private DoubleArray() { } | ||
|
|
||
| /** | ||
| * Returns a NEW {@link double[]} that is the cumulative sum of the passed array | ||
| * @param xs array to cumulatively sum | ||
| * @return new double[] | ||
| */ | ||
| public static double[] cumulativeSum(double[] xs) { | ||
|
There was a problem hiding this comment. Could you add a method comment? |
||
| double[] sum = new double[xs.length]; | ||
| sum[0] = xs[0]; | ||
| for (int i = 1; i < xs.length; i++) { | ||
| sum[i] = sum[i - 1] + xs[i]; | ||
| } | ||
| return sum; | ||
| } | ||
|
|
||
| /** | ||
| * Mutably divides each element in `xs` by the value `v` | ||
| * @param xs array to mutate with the divisor | ||
| * @param v The divisor, must not be 0.0, Infinite, or NaN. | ||
| */ | ||
| public static void divMut(double[] xs, double v) { | ||
|
There was a problem hiding this comment. Could you add a method comment? |
||
| if (v == 0.0 || Numbers.isValidDouble(v) == false) { | ||
| throw new IllegalArgumentException("unable to divide by [" + v + "] as it results in undefined behavior"); | ||
| } | ||
| for (int i = 0; i < xs.length; i++) { | ||
| xs[i] /= v; | ||
| } | ||
| } | ||
| } | ||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it is worth expanding this somewhat. For example, I think it is worth making clear that this is computing the K-S p-value between Y and E[Y | c] where c is the "event" selected by the sibling aggregation. The natural use case is when the sibling aggregation is a terms agg in which case it computes the K-S between Y and the restriction of Y to each term. This also means that one can't use any old values for fractions they have to match the actual proportion of the docs in each range bucket for Y.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I would propose something like the following:
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(I know you have some discussion of this later on, but I think this is crucial to understanding how this is meant to work for someone who knows what the K-S test actually does so worth including upfront.)