elasticsearch/docs/java-api/aggregations/metrics/scripted-metric-aggregation.asciidoc

[[java-aggs-metrics-scripted-metric]]
==== Scripted Metric Aggregation

Here is how you can use
{ref}/search-aggregations-metrics-scripted-metric-aggregation.html[Scripted Metric Aggregation]
with Java API.

===== Prepare aggregation request

Here is an example on how to create the aggregation request:

[source,java]
--------------------------------------------------
ScriptedMetricAggregationBuilder aggregation = AggregationBuilders
        .scriptedMetric("agg")
        .initScript(new Script("_agg['heights'] = []"))
        .mapScript(new Script("if (doc['gender'].value == \"male\") " +
                "{ _agg.heights.add(doc['height'].value) } " +
                "else " +
                "{ _agg.heights.add(-1 * doc['height'].value) }"));
--------------------------------------------------

You can also specify a `combine` script which will be executed on each shard:

[source,java]
--------------------------------------------------
ScriptedMetricAggregationBuilder aggregation =
        AggregationBuilders
                .scriptedMetric("agg")
                .initScript(new Script("_agg['heights'] = []"))
                .mapScript(new Script("if (doc['gender'].value == \"male\") " +
                        "{ _agg.heights.add(doc['height'].value) } " +
                        "else " +
                        "{ _agg.heights.add(-1 * doc['height'].value) }"))
                .combineScript(new Script("heights_sum = 0; for (t in _agg.heights) { heights_sum += t }; return heights_sum"));
--------------------------------------------------

You can also specify a `reduce` script which will be executed on the node which gets the request:

[source,java]
--------------------------------------------------
ScriptedMetricAggregationBuilder aggregation =
        AggregationBuilders
                .scriptedMetric("agg")
                .initScript(new Script("_agg['heights'] = []"))
                .mapScript(new Script("if (doc['gender'].value == \"male\") " +
                        "{ _agg.heights.add(doc['height'].value) } " +
                        "else " +
                        "{ _agg.heights.add(-1 * doc['height'].value) }"))
                .combineScript(new Script("heights_sum = 0; for (t in _agg.heights) { heights_sum += t }; return heights_sum"))
                .reduceScript(new Script("heights_sum = 0; for (a in _aggs) { heights_sum += a }; return heights_sum"));
--------------------------------------------------


===== Use aggregation response

Import Aggregation definition classes:

[source,java]
--------------------------------------------------
import org.elasticsearch.search.aggregations.bucket.terms.Terms;
import org.elasticsearch.search.aggregations.metrics.tophits.TopHits;
--------------------------------------------------

[source,java]
--------------------------------------------------
// sr is here your SearchResponse object
ScriptedMetric agg = sr.getAggregations().get("agg");
Object scriptedResult = agg.aggregation();
logger.info("scriptedResult [{}]", scriptedResult);
--------------------------------------------------

Note that the result depends on the script you built.
For the first example, this will basically produce:

[source,text]
--------------------------------------------------
scriptedResult object [ArrayList]
scriptedResult [ {
"heights" : [ 1.122218480146643, -1.8148918111233887, -1.7626731575142909, ... ]
}, {
"heights" : [ -0.8046067304119863, -2.0785486707864553, -1.9183567430207953, ... ]
}, {
"heights" : [ 2.092635728868694, 1.5697545960886536, 1.8826954461968808, ... ]
}, {
"heights" : [ -2.1863201099468403, 1.6328549117346856, -1.7078288405893842, ... ]
}, {
"heights" : [ 1.6043904836424177, -2.0736538674414025, 0.9898266674373053, ... ]
} ]
--------------------------------------------------

The second example will produce:

[source,text]
--------------------------------------------------
scriptedResult object [ArrayList]
scriptedResult [-41.279615707402876,
                -60.88007362339038,
                38.823270659734256,
                14.840192739445632,
                11.300902755741326]
--------------------------------------------------

The last example will produce:

[source,text]
--------------------------------------------------
scriptedResult object [Double]
scriptedResult [2.171917696507009]
--------------------------------------------------