honeymoose
/
OpenSearch
mirror of https://github.com/honeymoose/OpenSearch.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

[ML] model_debug_config: add enabled and remove bounds_percentile (elastic/x-pack-elasticsearch#627) 

This allows it to be disabled/enabled via job updates.
It also simplifies it by removing bounds_percentile as it currently
rarely makes sense to set the bounds to another value than 95.0.

Original commit: elastic/x-pack-elasticsearch@c27fce2d86
This commit is contained in:
Dimitris Athanasiou 2017-02-23 14:11:23 +00:00 committed by GitHub
parent 33b10970ee
commit 412d5e35b1
14 changed files with 82 additions and 115 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
plugin/src/main/java/org/elasticsearch/xpack/ml/job/config/ModelDebugConfig.java
View File

@ -11,57 +11,56 @@ import org.elasticsearch.common.io.stream.StreamInput;
import org.elasticsearch.common.io.stream.StreamOutput;
import org.elasticsearch.common.io.stream.Writeable;
import org.elasticsearch.common.xcontent.ConstructingObjectParser;
import org.elasticsearch.common.xcontent.ObjectParser.ValueType;
import org.elasticsearch.common.xcontent.XContentBuilder;
import org.elasticsearch.xpack.ml.job.messages.Messages;
import java.io.IOException;
import java.util.Locale;
import java.util.Objects;
public class ModelDebugConfig extends ToXContentToBytes implements Writeable {
private static final double MAX_PERCENTILE = 100.0;
private static final ParseField TYPE_FIELD = new ParseField("model_debug_config");
public static final ParseField BOUNDS_PERCENTILE_FIELD = new ParseField("bounds_percentile");
private static final ParseField ENABLED_FIELD = new ParseField("enabled");
public static final ParseField TERMS_FIELD = new ParseField("terms");
public static final ConstructingObjectParser<ModelDebugConfig, Void> PARSER = new ConstructingObjectParser<>(
TYPE_FIELD.getPreferredName(), a -> new ModelDebugConfig((Double) a[0], (String) a[1]));
TYPE_FIELD.getPreferredName(), a -> new ModelDebugConfig((boolean) a[0], (String) a[1]));
static {
PARSER.declareDouble(ConstructingObjectParser.constructorArg(), BOUNDS_PERCENTILE_FIELD);
PARSER.declareBoolean(ConstructingObjectParser.constructorArg(), ENABLED_FIELD);
PARSER.declareString(ConstructingObjectParser.optionalConstructorArg(), TERMS_FIELD);
}
private final double boundsPercentile;
private final boolean enabled;
private final String terms;
public ModelDebugConfig(double boundsPercentile, String terms) {
if (boundsPercentile < 0.0 || boundsPercentile > MAX_PERCENTILE) {
String msg = Messages.getMessage(Messages.JOB_CONFIG_MODEL_DEBUG_CONFIG_INVALID_BOUNDS_PERCENTILE);
throw new IllegalArgumentException(msg);
}
this.boundsPercentile = boundsPercentile;
public ModelDebugConfig() {
this(true, null);
}
public ModelDebugConfig(boolean enabled) {
this(false, null);
}
public ModelDebugConfig(boolean enabled, String terms) {
this.enabled = enabled;
this.terms = terms;
}
public ModelDebugConfig(StreamInput in) throws IOException {
boundsPercentile = in.readDouble();
enabled = in.readBoolean();
terms = in.readOptionalString();
}
@Override
public void writeTo(StreamOutput out) throws IOException {
out.writeDouble(boundsPercentile);
out.writeBoolean(enabled);
out.writeOptionalString(terms);
}
@Override
public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {
builder.startObject();
builder.field(BOUNDS_PERCENTILE_FIELD.getPreferredName(), boundsPercentile);
builder.field(ENABLED_FIELD.getPreferredName(), enabled);
if (terms != null) {
builder.field(TERMS_FIELD.getPreferredName(), terms);
}
@ -69,8 +68,8 @@ public class ModelDebugConfig extends ToXContentToBytes implements Writeable {
return builder;
}
public double getBoundsPercentile() {
return this.boundsPercentile;
public boolean isEnabled() {
return enabled;
}
public String getTerms() {
@ -88,11 +87,11 @@ public class ModelDebugConfig extends ToXContentToBytes implements Writeable {
}
ModelDebugConfig that = (ModelDebugConfig) other;
return Objects.equals(this.boundsPercentile, that.boundsPercentile) && Objects.equals(this.terms, that.terms);
return this.enabled == that.enabled && Objects.equals(this.terms, that.terms);
}
@Override
public int hashCode() {
return Objects.hash(boundsPercentile, terms);
return Objects.hash(enabled, terms);
}
}

2
plugin/src/main/java/org/elasticsearch/xpack/ml/job/messages/Messages.java
View File

@ -105,8 +105,6 @@ public final class Messages {
public static final String JOB_CONFIG_INVALID_TIMEFORMAT = "job.config.invalid.timeformat";
public static final String JOB_CONFIG_FUNCTION_INCOMPATIBLE_PRESUMMARIZED = "job.config.function.incompatible.presummarized";
public static final String JOB_CONFIG_MISSING_ANALYSISCONFIG = "job.config.missing.analysisconfig";
public static final String JOB_CONFIG_MODEL_DEBUG_CONFIG_INVALID_BOUNDS_PERCENTILE = "job.config.model.debug.config.invalid.bounds."
+ "percentile";
public static final String JOB_CONFIG_FIELD_VALUE_TOO_LOW = "job.config.field.value.too.low";
public static final String JOB_CONFIG_NO_ANALYSIS_FIELD = "job.config.no.analysis.field";
public static final String JOB_CONFIG_NO_ANALYSIS_FIELD_NOT_COUNT = "job.config.no.analysis.field.not.count";

5
plugin/src/main/java/org/elasticsearch/xpack/ml/job/process/autodetect/writer/ModelDebugConfigWriter.java
View File

@ -16,6 +16,9 @@ import static org.elasticsearch.xpack.ml.job.process.autodetect.writer.WriterCon
public class ModelDebugConfigWriter {
private static final double BOUNDS_PERCENTILE_DEFAULT = 95.0;
private static final double BOUNDS_PERCENTILE_DISABLE_VALUE = -1.0;
private final ModelDebugConfig modelDebugConfig;
private final Writer writer;
@ -29,7 +32,7 @@ public class ModelDebugConfigWriter {
contents.append("boundspercentile")
.append(EQUALS)
.append(modelDebugConfig.getBoundsPercentile())
.append(modelDebugConfig.isEnabled() ? BOUNDS_PERCENTILE_DEFAULT : BOUNDS_PERCENTILE_DISABLE_VALUE)
.append(NEW_LINE);
String terms = modelDebugConfig.getTerms();

1
plugin/src/main/resources/org/elasticsearch/xpack/ml/job/messages/ml_messages.properties
View File

@ -76,7 +76,6 @@ job.config.id.too.long = The job id cannot contain more than {0,number,integer}
job.config.invalid.fieldname.chars = Invalid field name ''{0}''. Field names including over, by and partition fields cannot contain any of these characters: {1}
job.config.invalid.timeformat = Invalid Time format string ''{0}''
job.config.missing.analysisconfig = An analysis_config must be set
job.config.model.debug.config.invalid.bounds.percentile = Invalid model_debug_config: bounds_percentile must be in the range [0, 100]
job.config.field.value.too.low = {0} cannot be less than {1,number}. Value = {2,number}
job.config.no.analysis.field = One of function, field_name, by_field_name or over_field_name must be set
job.config.no.analysis.field.not.count = Unless the function is 'count' one of field_name, by_field_name or over_field_name must be set

50
plugin/src/test/java/org/elasticsearch/xpack/ml/action/GetJobsActionResponseTests.java
View File

@ -7,19 +7,12 @@ package org.elasticsearch.xpack.ml.action;
import org.elasticsearch.xpack.ml.action.GetJobsAction.Response;
import org.elasticsearch.xpack.ml.action.util.QueryPage;
import org.elasticsearch.xpack.ml.job.config.AnalysisConfig;
import org.elasticsearch.xpack.ml.job.config.AnalysisLimits;
import org.elasticsearch.xpack.ml.job.config.DataDescription;
import org.elasticsearch.xpack.ml.job.config.Detector;
import org.elasticsearch.xpack.ml.job.config.Job;
import org.elasticsearch.xpack.ml.job.config.ModelDebugConfig;
import org.elasticsearch.xpack.ml.job.config.JobTests;
import org.elasticsearch.xpack.ml.support.AbstractStreamableTestCase;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Date;
import java.util.List;
import java.util.Map;
public class GetJobsActionResponseTests extends AbstractStreamableTestCase<GetJobsAction.Response> {
@ -30,46 +23,7 @@ public class GetJobsActionResponseTests extends AbstractStreamableTestCase<GetJo
int listSize = randomInt(10);
List<Job> jobList = new ArrayList<>(listSize);
for (int j = 0; j < listSize; j++) {
String jobId = "job" + j;
String description = randomBoolean() ? randomAsciiOfLength(100) : null;
Date createTime = new Date(randomNonNegativeLong());
Date finishedTime = randomBoolean() ? new Date(randomNonNegativeLong()) : null;
Date lastDataTime = randomBoolean() ? new Date(randomNonNegativeLong()) : null;
long timeout = randomNonNegativeLong();
AnalysisConfig.Builder analysisConfig = new AnalysisConfig.Builder(
Collections.singletonList(new Detector.Builder("metric", "some_field").build()));
AnalysisLimits analysisLimits = new AnalysisLimits(randomNonNegativeLong(), randomNonNegativeLong());
DataDescription.Builder dataDescription = new DataDescription.Builder();
ModelDebugConfig modelDebugConfig = randomBoolean() ? new ModelDebugConfig(randomDouble(), randomAsciiOfLength(10)) : null;
Long normalizationWindowDays = randomBoolean() ? Long.valueOf(randomIntBetween(0, 365)) : null;
Long backgroundPersistInterval = randomBoolean() ? Long.valueOf(randomIntBetween(3600, 86400)) : null;
Long modelSnapshotRetentionDays = randomBoolean() ? Long.valueOf(randomIntBetween(0, 365)) : null;
Long resultsRetentionDays = randomBoolean() ? Long.valueOf(randomIntBetween(0, 365)) : null;
Map<String, Object> customConfig = randomBoolean() ? Collections.singletonMap(randomAsciiOfLength(10), randomAsciiOfLength(10))
: null;
String modelSnapshotId = randomBoolean() ? randomAsciiOfLength(10) : null;
String indexName = "index" + j;
Job.Builder builder = new Job.Builder();
builder.setId(jobId);
builder.setDescription(description);
builder.setCreateTime(createTime);
builder.setFinishedTime(finishedTime);
builder.setLastDataTime(lastDataTime);
builder.setAnalysisConfig(analysisConfig);
builder.setAnalysisLimits(analysisLimits);
builder.setDataDescription(dataDescription);
builder.setModelDebugConfig(modelDebugConfig);
builder.setRenormalizationWindowDays(normalizationWindowDays);
builder.setBackgroundPersistInterval(backgroundPersistInterval);
builder.setModelSnapshotRetentionDays(modelSnapshotRetentionDays);
builder.setResultsRetentionDays(resultsRetentionDays);
builder.setCustomSettings(customConfig);
builder.setModelSnapshotId(modelSnapshotId);
builder.setResultsIndexName(indexName);
builder.setDeleted(randomBoolean());
Job job = builder.build();
jobList.add(job);
jobList.add(JobTests.createRandomizedJob());
}
result = new Response(new QueryPage<>(jobList, jobList.size(), Job.RESULTS_FIELD));

9
plugin/src/test/java/org/elasticsearch/xpack/ml/action/UpdateProcessActionRequestTests.java
View File

@ -5,11 +5,9 @@
*/
package org.elasticsearch.xpack.ml.action;
import org.elasticsearch.common.xcontent.XContentParser;
import org.elasticsearch.xpack.ml.job.config.JobUpdate;
import org.elasticsearch.xpack.ml.job.config.ModelDebugConfig;
import org.elasticsearch.xpack.ml.support.AbstractStreamableTestCase;
import org.elasticsearch.xpack.ml.support.AbstractStreamableXContentTestCase;
import java.util.List;
@ -20,11 +18,14 @@ public class UpdateProcessActionRequestTests extends AbstractStreamableTestCase<
protected UpdateProcessAction.Request createTestInstance() {
ModelDebugConfig config = null;
if (randomBoolean()) {
config = new ModelDebugConfig(5.0, "debug,config");
config = new ModelDebugConfig(randomBoolean(), randomAsciiOfLength(10));
}
List<JobUpdate.DetectorUpdate> updates = null;
if (randomBoolean()) {
int detectorUpdateCount = randomIntBetween(0, 5);
for (int i = 0; i < detectorUpdateCount; i++) {
new JobUpdate.DetectorUpdate(randomInt(), randomAsciiOfLength(10), null);
}
}
return new UpdateProcessAction.Request(randomAsciiOfLength(10), config, updates);
}

2
plugin/src/test/java/org/elasticsearch/xpack/ml/job/config/JobTests.java
View File

@ -397,7 +397,7 @@ public class JobTests extends AbstractSerializingTestCase<Job> {
outputs = new String[] {ac.getDetectors().get(0).getFieldName()};
}
if (randomBoolean()) {
builder.setModelDebugConfig(new ModelDebugConfig(randomDouble(), randomAsciiOfLength(10)));
builder.setModelDebugConfig(new ModelDebugConfig(randomBoolean(), randomAsciiOfLength(10)));
}
if (randomBoolean()) {
builder.setRenormalizationWindowDays(randomNonNegativeLong());

6
plugin/src/test/java/org/elasticsearch/xpack/ml/job/config/JobUpdateTests.java
View File

@ -48,7 +48,7 @@ public class JobUpdateTests extends AbstractSerializingTestCase<JobUpdate> {
update.setDetectorUpdates(detectorUpdates);
}
if (randomBoolean()) {
update.setModelDebugConfig(new ModelDebugConfig(randomDouble(), randomAsciiOfLength(10)));
update.setModelDebugConfig(new ModelDebugConfig(randomBoolean(), randomAsciiOfLength(10)));
}
if (randomBoolean()) {
update.setAnalysisLimits(new AnalysisLimits(randomNonNegativeLong(), randomNonNegativeLong()));
@ -99,7 +99,7 @@ public class JobUpdateTests extends AbstractSerializingTestCase<JobUpdate> {
new RuleCondition(RuleConditionType.NUMERICAL_ACTUAL, null, null, new Condition(Operator.GT, "5"), null))));
detectorUpdates.add(new JobUpdate.DetectorUpdate(1, "description-2", detectionRules2));
ModelDebugConfig modelDebugConfig = new ModelDebugConfig(randomDouble(), randomAsciiOfLength(10));
ModelDebugConfig modelDebugConfig = new ModelDebugConfig(randomBoolean(), randomAsciiOfLength(10));
AnalysisLimits analysisLimits = new AnalysisLimits(randomNonNegativeLong(), randomNonNegativeLong());
List<String> categorizationFilters = Arrays.asList(generateRandomStringArray(10, 10, false));
Map<String, Object> customSettings = Collections.singletonMap(randomAsciiOfLength(10), randomAsciiOfLength(10));
@ -153,7 +153,7 @@ public class JobUpdateTests extends AbstractSerializingTestCase<JobUpdate> {
public void testIsAutodetectProcessUpdate() {
JobUpdate update = new JobUpdate.Builder().build();
assertFalse(update.isAutodetectProcessUpdate());
update = new JobUpdate.Builder().setModelDebugConfig(new ModelDebugConfig(1.0, "ff")).build();
update = new JobUpdate.Builder().setModelDebugConfig(new ModelDebugConfig(true, "ff")).build();
assertTrue(update.isAutodetectProcessUpdate());
update = new JobUpdate.Builder().setDetectorUpdates(Arrays.asList(mock(JobUpdate.DetectorUpdate.class))).build();
assertTrue(update.isAutodetectProcessUpdate());

23
plugin/src/test/java/org/elasticsearch/xpack/ml/job/config/ModelDebugConfigTests.java
View File

@ -7,30 +7,21 @@ package org.elasticsearch.xpack.ml.job.config;
import org.elasticsearch.common.io.stream.Writeable.Reader;
import org.elasticsearch.common.xcontent.XContentParser;
import org.elasticsearch.test.ESTestCase;
import org.elasticsearch.xpack.ml.job.messages.Messages;
import org.elasticsearch.xpack.ml.support.AbstractSerializingTestCase;
import static org.hamcrest.Matchers.is;
import static org.hamcrest.Matchers.nullValue;
public class ModelDebugConfigTests extends AbstractSerializingTestCase<ModelDebugConfig> {
public void testVerify_GivenBoundPercentileLessThanZero() {
IllegalArgumentException e = ESTestCase.expectThrows(IllegalArgumentException.class, () -> new ModelDebugConfig(-1.0, ""));
assertEquals(Messages.getMessage(Messages.JOB_CONFIG_MODEL_DEBUG_CONFIG_INVALID_BOUNDS_PERCENTILE, ""), e.getMessage());
}
public void testVerify_GivenBoundPercentileGreaterThan100() {
IllegalArgumentException e = ESTestCase.expectThrows(IllegalArgumentException.class, () -> new ModelDebugConfig(100.1, ""));
assertEquals(Messages.getMessage(Messages.JOB_CONFIG_MODEL_DEBUG_CONFIG_INVALID_BOUNDS_PERCENTILE, ""), e.getMessage());
}
public void testVerify_GivenValid() {
new ModelDebugConfig(93.0, "");
new ModelDebugConfig(93.0, "foo,bar");
public void testConstructorDefaults() {
assertThat(new ModelDebugConfig().isEnabled(), is(true));
assertThat(new ModelDebugConfig().getTerms(), is(nullValue()));
}
@Override
protected ModelDebugConfig createTestInstance() {
return new ModelDebugConfig(randomDouble(), randomAsciiOfLengthBetween(1, 30));
return new ModelDebugConfig(randomBoolean(), randomAsciiOfLengthBetween(1, 30));
}
@Override

2
plugin/src/test/java/org/elasticsearch/xpack/ml/job/process/autodetect/AutodetectCommunicatorTests.java
View File

@ -56,7 +56,7 @@ public class AutodetectCommunicatorTests extends ESTestCase {
public void tesWriteUpdateModelDebugMessage() throws IOException {
AutodetectProcess process = mockAutodetectProcessWithOutputStream();
try (AutodetectCommunicator communicator = createAutodetectCommunicator(process, mock(AutoDetectResultProcessor.class))) {
ModelDebugConfig config = new ModelDebugConfig(10.0, "apple,peach");
ModelDebugConfig config = new ModelDebugConfig();
communicator.writeUpdateModelDebugMessage(config);
Mockito.verify(process).writeUpdateModelDebugMessage(config);
}

2
plugin/src/test/java/org/elasticsearch/xpack/ml/job/process/autodetect/NativeAutodetectProcessTests.java
View File

@ -145,7 +145,7 @@ public class NativeAutodetectProcessTests extends ESTestCase {
new AutodetectResultsParser(Settings.EMPTY))) {
process.start(executorService, mock(StateProcessor.class), mock(InputStream.class));
process.writeUpdateModelDebugMessage(new ModelDebugConfig(1.0, "term,s"));
process.writeUpdateModelDebugMessage(new ModelDebugConfig());
process.flushStream();
String message = new String(bos.toByteArray(), StandardCharsets.UTF_8);

4
plugin/src/test/java/org/elasticsearch/xpack/ml/job/process/autodetect/writer/ControlMsgToProcessWriterTests.java
View File

@ -151,12 +151,12 @@ public class ControlMsgToProcessWriterTests extends ESTestCase {
public void testWriteUpdateModelDebugMessage() throws IOException {
ControlMsgToProcessWriter writer = new ControlMsgToProcessWriter(lengthEncodedWriter, 2);
writer.writeUpdateModelDebugMessage(new ModelDebugConfig(10.0, "foo,bar"));
writer.writeUpdateModelDebugMessage(new ModelDebugConfig(true, "foo,bar"));
InOrder inOrder = inOrder(lengthEncodedWriter);
inOrder.verify(lengthEncodedWriter).writeNumFields(4);
inOrder.verify(lengthEncodedWriter, times(3)).writeField("");
inOrder.verify(lengthEncodedWriter).writeField("u[modelDebugConfig]\nboundspercentile = 10.0\nterms = foo,bar\n");
inOrder.verify(lengthEncodedWriter).writeField("u[modelDebugConfig]\nboundspercentile = 95.0\nterms = foo,bar\n");
verifyNoMoreInteractions(lengthEncodedWriter);
}

39
plugin/src/test/java/org/elasticsearch/xpack/ml/job/process/autodetect/writer/ModelDebugConfigWriterTests.java
View File

@ -5,17 +5,17 @@
*/
package org.elasticsearch.xpack.ml.job.process.autodetect.writer;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.verifyNoMoreInteractions;
import org.elasticsearch.test.ESTestCase;
import org.elasticsearch.xpack.ml.job.config.ModelDebugConfig;
import org.junit.After;
import org.junit.Before;
import org.mockito.Mockito;
import java.io.IOException;
import java.io.OutputStreamWriter;
import org.elasticsearch.test.ESTestCase;
import org.junit.After;
import org.junit.Before;
import org.mockito.Mockito;
import org.elasticsearch.xpack.ml.job.config.ModelDebugConfig;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.verifyNoMoreInteractions;
public class ModelDebugConfigWriterTests extends ESTestCase {
private OutputStreamWriter writer;
@ -29,12 +29,31 @@ public class ModelDebugConfigWriterTests extends ESTestCase {
public void verifyNoMoreWriterInteractions() {
verifyNoMoreInteractions(writer);
}
public void testWrite_GivenFullConfig() throws IOException {
ModelDebugConfig modelDebugConfig = new ModelDebugConfig(65.0, "foo,bar");
public void testWrite_GivenEnabledConfigWithoutTerms() throws IOException {
ModelDebugConfig modelDebugConfig = new ModelDebugConfig();
ModelDebugConfigWriter writer = new ModelDebugConfigWriter(modelDebugConfig, this.writer);
writer.write();
verify(this.writer).write("boundspercentile = 65.0\nterms = foo,bar\n");
verify(this.writer).write("boundspercentile = 95.0\nterms = \n");
}
public void testWrite_GivenEnabledConfigWithTerms() throws IOException {
ModelDebugConfig modelDebugConfig = new ModelDebugConfig(true, "foo,bar");
ModelDebugConfigWriter writer = new ModelDebugConfigWriter(modelDebugConfig, this.writer);
writer.write();
verify(this.writer).write("boundspercentile = 95.0\nterms = foo,bar\n");
}
public void testWrite_GivenDisabledConfigWithTerms() throws IOException {
ModelDebugConfig modelDebugConfig = new ModelDebugConfig(false, "foo,bar");
ModelDebugConfigWriter writer = new ModelDebugConfigWriter(modelDebugConfig, this.writer);
writer.write();
verify(this.writer).write("boundspercentile = -1.0\nterms = foo,bar\n");
}
}

9
plugin/src/test/resources/rest-api-spec/test/ml/jobs_crud.yaml
View File

@ -198,7 +198,8 @@
"field_delimiter":","
},
"model_debug_config": {
"bounds_percentile": 95.0
"enabled": true,
"terms": "foo,bar"
},
"analysis_limits": {
"model_memory_limit": 10
@ -229,7 +230,8 @@
"condition": {"operator": "gt", "value": "10" } } ] } },
{"index": 1, "description": "updated description"}],
"model_debug_config": {
"bounds_percentile": 99.0
"enabled": false,
"terms": "foobar"
},
"analysis_limits": {
"model_memory_limit": 20
@ -245,7 +247,8 @@
}
- match: { job_id: "to-update" }
- match: { description: "Post update description" }
- match: { model_debug_config.bounds_percentile: 99.0 }
- match: { model_debug_config.enabled: false }
- match: { model_debug_config.terms: "foobar" }
- match: { analysis_limits.model_memory_limit: 20 }
- match: { analysis_config.categorization_filters: ["cat3.*"] }
- match: { analysis_config.detectors.0.detector_rules.0.target_field_name: "airline" }