OpenSearch/docs/java-rest/high-level/ml/estimate-model-memory.asciidoc

43 lines
1.4 KiB
Plaintext
Raw Normal View History

--
:api: estimate-model-memory
:request: EstimateModelMemoryRequest
:response: EstimateModelMemoryResponse
--
[role="xpack"]
[id="{upid}-{api}"]
=== Estimate {anomaly-job} model memory API
Estimate the model memory an analysis config is likely to need for
the given cardinality of the fields it references.
[id="{upid}-{api}-request"]
==== Estimate {anomaly-job} model memory request
A +{request}+ can be set up as follows:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-request]
--------------------------------------------------
<1> Pass an `AnalysisConfig` to the constructor.
<2> For any `by_field_name`, `over_field_name` or
`partition_field_name` fields referenced by the
detectors, supply overall cardinality estimates
in a `Map`.
<3> For any `influencers`, supply a `Map` containing
estimates of the highest cardinality expected in
any single bucket.
include::../execution.asciidoc[]
[id="{upid}-{api}-response"]
==== Estimate {anomaly-job} model memory response
The returned +{response}+ contains the model memory estimate:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-response]
--------------------------------------------------
<1> The model memory estimate.