Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg #29594

polyfractal · 2018-04-18T20:03:29Z

This pipeline aggregation gives the user the ability to script functions that "move" across a window
of data, instead of single data points. Through custom script contexts, we also expose a number of convenience methods:

windowMax()
windowMin()
windowSum()
simpleMovAvg()
linearMovAvg()
ewmaMovAvg()
holtMovAvg()
holtWintersMovAvg()

Since we can easily expose moving average functionality through custom script contexts, I would like to deprecate the moving_avg agg at the same time if there isn't objection. Thoughts?

At the moment, this agg doesn't support "predictions" like the old agg... but I think that is a good thing. The "prediction" functionality is just super basic extrapolation, is fairly confusing for users and of questionable value. It also means we can eventually remove the simulated annealing optimizer, since that was mainly there to help minimize the parameters for holt-winters prediction.

I gutted the math portions of the moving avg models and consolidated them in this PR. There's still a lot of boilerplate cruft from the old models left, but that's just the way it is. In the process I de-generified the models to make it work with MovFn, and generics weren't needed anyway.

This is still a WIP, needs docs, etc. @rjernst would you mind taking a look at the scripting context stuff to make sure it looks ok? I'll track down a search/aggs team member for the rest of the review when it's ready.

This is a reboot of #25137

This pipeline aggregation gives the user the ability to script functions that "move" across a window of data, instead of single data points. It is the scripted version of MovingAvg pipeline agg. Through custom script contexts, we expose a number of convenience methods: - windowMax - windowMin - windowSum - simpleMovAvg - linearMovAvg - ewmaMovAvg - holtMovAvg - holtWintersMovAvg The user can also define any arbitrary logic via their own scripting, or combine with the above methods.

No need to c/p the code, we can basically share the movavg models with an easy path towards deprecation in the future.

elasticmachine · 2018-04-18T20:03:31Z

Pinging @elastic/es-search-aggs

colings86

@polyfractal I took a first pass on this and left some comments but I like the approach

colings86 · 2018-04-26T09:09:13Z

modules/lang-painless/src/test/resources/rest-api-spec/test/painless/70_mov_fn_agg.yml

+                  moving_fn:
+                    buckets_path: "the_avg"
+                    window: 3
+                    script: "MovingFunctions.windowMax(values)"


I wonder if we should test that the other functions are exposed correctly too?

colings86 · 2018-04-26T09:10:38Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/BucketHelpers.java

@@ -79,7 +79,7 @@ public static GapPolicy parse(String text, XContentLocation tokenLocation) {
                for (GapPolicy policy : values()) {
                    validNames.add(policy.getName());
                }
-                throw new ParsingException(tokenLocation, "Invalid gap policy: [" + text + "], accepted values: " + validNames);
+                throw new IllegalArgumentException("Invalid gap policy: [" + text + "], accepted values: " + validNames);


Why are we removing the token location here, it seems like useful information to return to the user?

I thought it was incompatible with the new parser style (I didnt think you could get token info with the new way of writing parsers), but while writing this comment I noticed how BucketSort did it:

PARSER.declareField(BucketSortPipelineAggregationBuilder::gapPolicy, p -> { if (p.currentToken() == XContentParser.Token.VALUE_STRING) { return GapPolicy.parse(p.text().toLowerCase(Locale.ROOT), p.getTokenLocation()); } throw new IllegalArgumentException("Unsupported token [" + p.currentToken() + "]"); }

which I somehow missed before. I'll put it back to the way it was and using the above snippet :)

colings86 · 2018-04-26T09:15:40Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+    /**
+     * Find the maximum value in a window of values
+     */
+    public static double windowMax(Collection<Double> values) {


In the JavaDocs can we qualify the behaviour if there are no values for all these functions? e.g. for this one it will return Double.NaN

colings86 · 2018-04-26T09:17:03Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+     */
+    public static double linearMovAvg(Collection<Double> values) {
+        double avg = 0;
+        long totalWeight = 1;


is it correct that the total weight starts at 1 here? This means that if there are no values the function will return 0 which seems misleading?

Hmm, good point. OTOH, that's how the linear avg works today so it'd be "breaking" the current behavior (these functions were basically lifted directly from existing moving model code)

I think we should probably change it anyway, but wanted to raise that point.

colings86 · 2018-04-26T09:18:27Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+                avg = (v * alpha) + (avg * (1 - alpha));
+            }
+        }
+        return avg;


This function will also return 0 if there are no values, is this intended?

Ditto above

polyfractal · 2018-04-27T18:22:43Z

Review comments addressed, and added documentation + deprecation notices + more tests

colings86

I left some comments, mostly about documentation

colings86 · 2018-04-30T09:10:07Z

docs/reference/aggregations/pipeline/movfn-aggregation.asciidoc

+etc.
+
+This is conceptually very similar to the <<search-aggregations-pipeline-movavg-aggregation, Moving Average>> pipeline aggregation, except
+it provides more functionality and better flexibility.


more functionality and better flexibility -> I think we can just say more functionality here rather than repeating?

colings86 · 2018-04-30T09:11:10Z

docs/reference/aggregations/pipeline/movfn-aggregation.asciidoc

+|`buckets_path` |Path to the metric of interest (see <<buckets-path-syntax, `buckets_path` Syntax>> for more details |Required |
+|`window` |The size of window to "slide" across the histogram. |Optional |`5`
+|`script` |The script that should be executed on each window of data |Required |
+|`params` |User-defined extra parameters to be injected into the script |Optional |


This shouldn't be needed as the script field itself can handle params?

colings86 · 2018-04-30T09:13:05Z

docs/reference/aggregations/pipeline/movfn-aggregation.asciidoc

+// TEST[setup:sales]
+
+<1> A `date_histogram` named "my_date_histo" is constructed on the "timestamp" field, with one-day intervals
+<2> A `sum` metric is used to calculate the sum of a field.  This could be any metric (sum, min, max, etc)


I think we should say this could be any numeric metric as I don't think you could point a moving_fn` at, for example, a geo_centroid agg?

colings86 · 2018-04-30T09:13:39Z

docs/reference/aggregations/pipeline/movfn-aggregation.asciidoc

+<3> Finally, we specify a `moving_fn` aggregation which uses "the_sum" metric as its input.
+
+Moving averages are built by first specifying a `histogram` or `date_histogram` over a field.  You can then optionally
+add normal metrics, such as a `sum`, inside of that histogram.  Finally, the `moving_fn` is embedded inside the histogram.


normal -> numeric ?

colings86 · 2018-04-30T09:17:40Z

docs/reference/aggregations/pipeline/movfn-aggregation.asciidoc

+- `linearMovAvg()`
+- `ewmaMovAvg()`
+- `holtMovAvg()`
+- `holtWintersMovAvg()`


I wonder if it would be useful to add windowStdDev() to this list too as I have seen some requests for moving standard deviations too. However I imagine given that it would be used together with moving average, it would need the simple, linear, ewma, holt and holtWinters forms to make sense with moving average, so maybe its too complex to add it in this PR?

colings86 · 2018-04-30T09:36:15Z

.../src/main/java/org/elasticsearch/search/aggregations/pipeline/movavg/models/SimpleModel.java

@@ -84,7 +84,11 @@ public MovAvgModel clone() {

    @Override
    public double next(Collection<Double> values) {
-        return MovingFunctions.simpleMovAvg(values);
+        double avg = 0;


can we call this sum as its not actually the avg? 😉

colings86 · 2018-04-30T09:38:10Z

.../src/main/java/org/elasticsearch/search/aggregations/pipeline/movavg/models/LinearModel.java

@@ -86,7 +86,16 @@ public MovAvgModel clone() {

    @Override
    public double next(Collection<Double> values) {


Can we add a comment here explaining why this is duplicating functionality provided by MovingFunctions? I think someone could try to deduplicate the code without this

Oof, actually, this should be pulled back out. This was added when I thought it was a breaking change but it's not actually (since we don't allow null/NaN as a value because of gap_policy) in regular mov avg.

Thanks for noticing, I'll revert this to pointing at the static MovingFunctions version.

colings86 · 2018-04-30T09:39:47Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

     */
    public static double windowSum(Collection<Double> values) {
-        return values.stream().mapToDouble(Double::doubleValue).sum();
+        if (values.size() == 0) {
+            return 0.0;


From the docs change above it says:

If the window is empty, or all values are null/NaN, 0.0 is returned as the result.

This seems to not be the case here?

? If the size is zero, it returns 0.0. And in the mapToDouble(), we only use the value if it's non-null and non-NaN (otherwise we return 0.0 for that position).

There's a test to verify this: https://github.com/elastic/elasticsearch/pull/29594/files/83778999a9aec1e0170efba0d4857100a927a770..c93d065b6ffd4e37089b4e251ad784380b75a942#diff-a35897b4e9c0a693a4515cab553a6427R73

colings86 · 2018-04-30T09:42:15Z

...va/org/elasticsearch/search/aggregations/pipeline/movfn/MovFnPipelineAggregationBuilder.java

+    private final String bucketsPathString;
+    private String format = null;
+    private GapPolicy gapPolicy = GapPolicy.SKIP;
+    private int window;


Since this is optional we should set the default here

Doc error. Window isn't optional, I'll make the fix in the documentation.

colings86 · 2018-04-30T09:42:51Z

...va/org/elasticsearch/search/aggregations/pipeline/movfn/MovFnPipelineAggregationBuilder.java

+    };
+
+
+    public MovFnPipelineAggregationBuilder(String name, String bucketsPath, Script script, int window) {


Since the window is optional we should not provide it in the constructor here but rather with a setter method.

polyfractal · 2018-04-30T22:02:11Z

Added more details in docs about handling of NaN and null, and how it affects avg counts
Made window mandatory in docs
Added a standard deviation. I decided to sidestep the issue about what avg to use (and proliferation of a std dev to match each avg) by just making the user pass in the avg. The example shows how to combine one of the moving avg functions with std dev to get the desired behavior. Seemed only slightly less user-friendly, but a lot more flexible and less hassle to maintain
Fixed up a bunch of doc testing issues

rjernst

The script part looks pretty good. I left a couple suggestions.

rjernst · 2018-04-30T22:30:58Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovFnScript.java

+ * This class provides a custom script context for the Moving Function pipeline aggregation,
+ * so that we can expose a number of pre-baked moving functions like min, max, movavg, etc
+ */
+public abstract class MovFnScript {


Can the name not be so cryptic? MovingFunctionScript?

rjernst · 2018-04-30T22:40:32Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+     * If all values are missing/null/NaN, the return value will be NaN.
+     * The average is based on the count of non-null, non-NaN values.
+     */
+    public static double simpleMovAvg(Collection<Double> values) {


Maybe just call this unweightedAvg? Is the MovAvg suffix really necessary given this is within MovingFunctions? Ditto for the methods below.

rjernst · 2018-04-30T22:42:41Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+     * Find the maximum value in a window of values.
+     * If all values are missing/null/NaN, the return value will be NaN
+     */
+    public static double windowMax(Collection<Double> values) {


What do you think of using DoubleStream as the type here? I think it could save boxing in some cases?

I was ++ on this and got it all coded up, but realized this makes certain things impossible because streams can only be used once.

The standard deviation for example just accepts an avg and the values:

MovingFunctions.windowStdDev(values, MovingFunctions.simpleMovAvg(values))

Which won't work because the stream will get used up in the movavg, and throw an exception when we try to iterate on it again :(

polyfractal · 2018-05-02T17:32:45Z

More cleanup:

Renamed MovFnScript to MovingFunctionScript
Test fix for movAvg, needed to assertWarning in a few places
Discussed with @colings86 about naming and we decided to overhaul all the names, not just simpleMovAvg. The list is now:
- max()
- min()
- sum()
- stdDev()
- unweightedAvg()
- linearWeightedAvg()
- ewma()
- holt()
- holtWinters()
  Chose to keep ewma/holt/holtWinters instead of exponentionalWeighted/doubleExponentialWeighted/tripleExponentialWeighted for brevity :)

rjernst

I like the new names! I left a couple more suggestions.

rjernst · 2018-05-02T22:07:57Z

...src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctionScript.java

+    }
+
+    public static final String[] PARAMETERS = new String[] {"params", "values"};
+    public static final ScriptContext<Factory> CONTEXT = new ScriptContext<>("movfn", Factory.class);


I think this should be "moving-function" to parallel the class name? While this is unseen for users most of the time, it can be used when putting a stored script to compile against that context.

rjernst · 2018-05-02T22:15:05Z

server/src/main/java/org/elasticsearch/search/aggregations/pipeline/movfn/MovingFunctions.java

+     * Find the maximum value in a window of values.
+     * If all values are missing/null/NaN, the return value will be NaN
+     */
+    public static double max(Collection<Double> values) {


I have 2 concerns still:

Painless doesn't know anything about generics. So to a script, this means passing in any Collection (that is all painless will check for). What about double[]?

Collection is unordered in java, but the order matters (I think?) for some of these calculations (not max, but others). Using double[] would solve this, but if we must stick with boxed types, at least use List<Double>?

Note that doc['fieldname'] returns an object that implements List, but we could make this easier by adding an asArray() method to DoubleScriptDocValues?

I think double[] should be doable. 👍

I didn't do that originally to avoid unboxing the internal EvictingQueue into an array since it wouldn't be necessary most of the time. And I avoided specifying a Queue as the type since the user might want to use these functions outside of the values we provide (or in addition to).

But I didn't think about the unordered nature of Collection, that's definitely an issue. Lemme poke at it tomorrow with double[], I think that'll work fine.

Sidebar
Alternatively, I could replace the evicting queue with a primitive-based ring buffer. I don't think that's any better though. We'd still have to arrayCopy the relevant slice into the script since there are null/NaNs to deal with, and we probably shouldn't make the method signature some weird ring buffer class :)

Note that doc['fieldname'] returns an object that implements List, but we could make this easier by adding an asArray() method to DoubleScriptDocValues?

Ah interesting. I'm not sure it's needed... do we think users will plop a field's value into the functions instead of the values? I was thinking users might tweak the values before a function, or write their own based on the values... but not use an entire field in place of the provided values. Thoughts?

Forget what I said about ScriptDocValues. I was thinking this was like scripted metric agg, which uses a SearchScript for the map script, but you did this correctly (no getDoc()). So no need to do anything with the doc values accessors.

polyfractal · 2018-05-03T15:44:35Z

@rjernst swapped over to double[]. As a side benefit, it makes things a lot easier for the user in the script (values[0] just seems more ergonomic than working with collections/lists) :)

These will be bumped to `6.3.99` after the feature has been backported to 6.4

polyfractal · 2018-05-08T21:01:40Z

@colings86 @rjernst did either of you want another review pass on this?

rjernst

LGTM

colings86

LGTM too

Test was previously muted which hid the failure. The holt model in the test needed updating like the real model (was using incorrect starting count)

…ngs-to-true * elastic/master: (34 commits) Test: increase search logging for LicensingTests Adjust serialization version in IndicesOptions [TEST] Fix compilation Remove version argument in RangeFieldType (elastic#30411) Remove unused DirectoryUtils class. (elastic#30582) Mitigate date histogram slowdowns with non-fixed timezones. (elastic#30534) Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg (elastic#29594) Removes AwaitsFix on IndicesOptionsTests Template upgrades should happen in a system context (elastic#30621) Fix bug in BucketMetrics path traversal (elastic#30632) Fixes IndiceOptionsTests to serialise correctly (elastic#30644) S3 repo plugin populate SettingsFilter (elastic#30652) mute IndicesOptionsTests.testSerialization Rest High Level client: Add List Tasks (elastic#29546) Refactors ClientHelper to combine header logic (elastic#30620) [ML] Wait for ML indices in rolling upgrade tests (elastic#30615) Watcher: Ensure secrets integration tests also run triggered watch (elastic#30478) Move allocation awareness attributes to list setting (elastic#30626) [Docs] Update code snippet in has-child-query.asciidoc (elastic#30510) Replace custom reloadable Key/TrustManager (elastic#30509) ...

* es/master: (74 commits) Preserve REST client auth despite 401 response (#30558) [test] packaging: add windows boxes (#30402) Make xpack modules instead of a meta plugin (#30589) Mute ShrinkIndexIT [ML] DeleteExpiredDataAction should use client with origin (#30646) Reindex: Fixed typo in assertion failure message (#30619) [DOCS] Fixes list of unconverted snippets in build.gradle [DOCS] Reorganizes RBAC documentation SQL: Remove dependency for server's version from JDBC driver (#30631) Test: increase search logging for LicensingTests Adjust serialization version in IndicesOptions [TEST] Fix compilation Remove version argument in RangeFieldType (#30411) Remove unused DirectoryUtils class. (#30582) Mitigate date histogram slowdowns with non-fixed timezones. (#30534) Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg (#29594) Removes AwaitsFix on IndicesOptionsTests Template upgrades should happen in a system context (#30621) Fix bug in BucketMetrics path traversal (#30632) Fixes IndiceOptionsTests to serialise correctly (#30644) ...

* es/ccr: (75 commits) Preserve REST client auth despite 401 response (elastic#30558) [test] packaging: add windows boxes (elastic#30402) Make xpack modules instead of a meta plugin (elastic#30589) Mute ShrinkIndexIT [ML] DeleteExpiredDataAction should use client with origin (elastic#30646) Reindex: Fixed typo in assertion failure message (elastic#30619) [DOCS] Fixes list of unconverted snippets in build.gradle [DOCS] Reorganizes RBAC documentation SQL: Remove dependency for server's version from JDBC driver (elastic#30631) Test: increase search logging for LicensingTests Adjust serialization version in IndicesOptions [TEST] Fix compilation Remove version argument in RangeFieldType (elastic#30411) Remove unused DirectoryUtils class. (elastic#30582) Mitigate date histogram slowdowns with non-fixed timezones. (elastic#30534) Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg (elastic#29594) Removes AwaitsFix on IndicesOptionsTests Template upgrades should happen in a system context (elastic#30621) Fix bug in BucketMetrics path traversal (elastic#30632) Fixes IndiceOptionsTests to serialise correctly (elastic#30644) ...

…lastic#29594) This pipeline aggregation gives the user the ability to script functions that "move" across a window of data, instead of single data points. It is the scripted version of MovingAvg pipeline agg. Through custom script contexts, we expose a number of convenience methods: - MovingFunctions.max() - MovingFunctions.min() - MovingFunctions.sum() - MovingFunctions.unweightedAvg() - MovingFunctions.linearWeightedAvg() - MovingFunctions.ewma() - MovingFunctions.holt() - MovingFunctions.holtWinters() - MovingFunctions.stdDev() The user can also define any arbitrary logic via their own scripting, or combine with the above methods.

See elastic/elasticsearch#29594

See elastic/elasticsearch#29594 (cherry picked from commit 6c4f69f)

Documents the removal of the `moving_avg` aggregation in the 8.0 breaking changes. Relates to #29594.

$polyfractal$

polyfractal added 2 commits April 18, 2018 19:29

$@polyfractal$

Consolidate movavg models into MovingFunctions

8377899

No need to c/p the code, we can basically share the movavg models with an easy path towards deprecation in the future.

$@polyfractal$ polyfractal added WIP :Analytics/Aggregations Aggregations labels Apr 18, 2018

colings86 added the >feature label Apr 24, 2018

colings86 reviewed Apr 26, 2018

View reviewed changes

polyfractal added 3 commits April 27, 2018 11:35

$@polyfractal$

Merge remote-tracking branch 'origin/master' into moving_fn_v2

8dbb601

$@polyfractal$

Review cleanup

9c9934a

$@polyfractal$

Add documentation and depreaction warnings

c93d065

$@polyfractal$ polyfractal added review and removed WIP labels Apr 27, 2018

colings86 requested changes Apr 30, 2018

View reviewed changes

$@polyfractal$

More review cleanup, add std dev, fix doc tests

0988e82

rjernst reviewed Apr 30, 2018

View reviewed changes

polyfractal added 3 commits May 2, 2018 16:30

$@polyfractal$

Review cleanup - renaming, test fixes

a9d3be4

$@polyfractal$

Merge remote-tracking branch 'origin/master' into moving_fn_v2

17b7408

$@polyfractal$

Test fix

635ca8f

rjernst reviewed May 2, 2018

View reviewed changes

polyfractal added 2 commits May 3, 2018 15:32

$@polyfractal$

Switch Collection to double[], change ScriptContext name

43eee62

$@polyfractal$

Merge remote-tracking branch 'origin/master' into moving_fn_v2

7747fab

polyfractal added 5 commits May 3, 2018 16:51

$@polyfractal$

[TEST] add warning header to rest test

08b915c

$@polyfractal$

Add mock movingfunction script

2b27b5d

$@polyfractal$

[DOCS] Update changelog

121b31d

$@polyfractal$

Merge remote-tracking branch 'origin/master' into moving_fn_v2

34346cc

$@polyfractal$

[DOCS] Fix conflict

a7cb005

polyfractal added 2 commits May 7, 2018 14:59

$@polyfractal$

Merge remote-tracking branch 'origin/master' into moving_fn_v2

505ee27

$@polyfractal$

Add version skips

25a226a

These will be bumped to `6.3.99` after the feature has been backported to 6.4

rjernst approved these changes May 9, 2018

View reviewed changes

colings86 approved these changes May 15, 2018

View reviewed changes

polyfractal added 2 commits May 15, 2018 17:50

$@polyfractal$

Merge remote-tracking branch 'origin' into moving_fn_v2

a5393e6

$@polyfractal$

Fix holt test

e7b35cf

Test was previously muted which hid the failure. The holt model in the test needed updating like the real model (was using incorrect starting count)

$@polyfractal$ polyfractal merged commit df853c4 into elastic:master May 16, 2018

$@polyfractal$ polyfractal added v7.0.0 v6.4.0 and removed review labels May 16, 2018

This was referenced May 16, 2018

MovingFunction Pipeline agg backport to 6.x #30658

Merged

Additional "moving" pipeline aggregations #17607

Closed

Mpdreamz mentioned this pull request Sep 21, 2018

[meta] 6.4.0 release elastic/elasticsearch-net#3397

Closed

89 tasks

Mpdreamz added a commit to elastic/elasticsearch-net that referenced this pull request Sep 24, 2018

Add support for moving function aggregate elastic/elasticsearch#29594

240ef0c

Mpdreamz mentioned this pull request Sep 24, 2018

Add support for moving function aggregation elastic/elasticsearch-net#3419

Merged

russcam pushed a commit to elastic/elasticsearch-net that referenced this pull request Oct 16, 2018

Add support for Moving Function aggregation (#3419)

6c4f69f

See elastic/elasticsearch#29594

russcam pushed a commit to elastic/elasticsearch-net that referenced this pull request Oct 26, 2018

Add support for Moving Function aggregation (#3419)

4e62181

See elastic/elasticsearch#29594 (cherry picked from commit 6c4f69f)

colings86 added v7.0.0-beta1 and removed v7.0.0 labels Feb 7, 2019

markov00 mentioned this pull request May 3, 2019

Replace Moving Average pipeline aggregation elastic/kibana#24702

Closed

2 tasks

jrodewig mentioned this pull request Oct 30, 2020

[DOCS] Redirect moving avg aggregation #64435

Merged

jrodewig mentioned this pull request Sep 20, 2021

[DOCS] Add breaking change for moving_avg #78018

Merged

jrodewig added a commit that referenced this pull request Sep 20, 2021

[DOCS] Add breaking change for moving_avg (#78018)

f3f11ab

Documents the removal of the `moving_avg` aggregation in the 8.0 breaking changes. Relates to #29594.

joegallo mentioned this pull request Jan 19, 2022

[Rest Api Compatibility] MovAvgPipelineAggregation #82828

Merged

		@@ -86,7 +86,16 @@ public MovAvgModel clone() {

		@Override
		public double next(Collection<Double> values) {

		};


		public MovFnPipelineAggregationBuilder(String name, String bucketsPath, Script script, int window) {

Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg #29594

Add a MovingFunction pipeline aggregation, deprecate MovingAvg agg #29594

Conversation

polyfractal commented Apr 18, 2018

elasticmachine commented Apr 18, 2018

colings86 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

colings86 Apr 26, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

polyfractal commented Apr 27, 2018

colings86 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

polyfractal commented Apr 30, 2018

rjernst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

polyfractal commented May 2, 2018

rjernst left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rjernst May 2, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rjernst May 2, 2018 • edited Loading

Choose a reason for hiding this comment

polyfractal commented May 3, 2018

polyfractal commented May 8, 2018

rjernst left a comment

Choose a reason for hiding this comment

colings86 left a comment

Choose a reason for hiding this comment

$@polyfractal$ polyfractal commented Apr 18, 2018

colings86 Apr 26, 2018 •

edited

Loading

rjernst May 2, 2018 •

edited

Loading

rjernst May 2, 2018 •

edited

Loading